Evans-Myron: Arbiter (Step 3)

Inputs reconciled: claims.md, facts.json, physicist.md, mathematician.md, contrarian.md. Tags: [read], [calc] (my own SymPy run), [web] URL, [memory] with confidence.

Source gap (repeated from Step 1)

The folder holds only two books: Evans, Eckardt, Crothers and Pendergast, Criticisms of the Einstein Field Equation (CEFE, 2010, 548 pp.) and Felker, The Evans Equations of Unified Field Theory (2005, 373 pp., a popularisation not written by Evans). There are no Evans papers (AIAS papers 63, 90, 93, 94, 122 are cited but absent), no peer-reviewed material, no device descriptions and no experimental data. (Updated after QE, see "QE revision" below: Crothers Secs 3.5-3.11, Ch.4 Secs 4.1-4.3 and Felker Chs 9-12 now have deep reads; the 62 Ch.4 tables were checked by their printed "o.k." lines and by sampled recomputation, not one by one; CEFE Sec. 2.8 rests on the extracted text plus QE's recomputation.) Originally: the 62 metric tables in CEFE Ch.4, most of Crothers's Ch.3 (including the gravitational-wave section) and most of Felker were not read by anyone. Every statement about replications, critics and referee behaviour rests on one secondary web source (Wikipedia, Einstein-Cartan-Evans theory) plus the web checks listed at the end. Conclusions that rest on [memory] alone are flagged below. To close the gap, obtain: AIAS papers 63, 94 (Coulomb resonance and Bedini application), 93 (Physica B submission and referee reports), the Hehl-Obukhov and Bruhn critiques, and any measured input/output data for the claimed devices.

Where the reviewers agreed and disagreed (overview)

All three agree that Evans's central geometric argument (Claim 1) fails, that the Ch.4 computer-algebra "test" measures Ricci, and that the spin-connection resonance in Sec. 2.6 depends on a complex-radius substitution and has no energy budget. They differ on (i) whether Claim 3 is "internally inconsistent" (physicist) or "not testable as stated" (mathematician, contrarian); (ii) whether the uncertainty-principle refutation is "refuted by calculation" (mathematician) or "tested, not supported" (physicist, contrarian); (iii) the label for the Evans wave equation and for "no Higgs"; (iv) how to word the Bedini/free-power claim.

Recomputed numbers (differences between reviewers)

Run in .venv, scratch /tmp/arb_em/a.py.

1. Ricci identification. FLRW metric with scale factor a(t): the full contraction of the Riemann tensor gives 00 component -3a''/a and spatial components (a a'' + 2a'^2) (in index-lowered form). These are plus or minus the Ricci tensor, as all three reviewers said. [calc] Hence Evans's R^{kappa mu}_{mu nu} (Eq. 4.8) is a Ricci contraction, and "nonzero for every matter solution" restates the Einstein equation. (The sign differs between mixed and lowered forms, which explains the apparent +3a''/a versus -3a''/a wording in the contrarian's note; it does not matter for the verdict.)
2. Radial Poisson equation with spin connection. For radial omega the equation is phi'' + (2/r + omega) phi' + (omega' + 2 omega/r) phi = -rho/eps0. With omega = -1/r the coefficients are 1/r and -1/r^2, matching Eq. 2.156; homogeneous solutions r and 1/r (indicial s^2 - 1 = 0), no oscillation. [calc] Physicist, mathematician and contrarian agree.
3. Eqs. 2.161-2.165. With omega = omega0^2 r - 4 beta ln r - 4/r the phi' coefficient is -4 beta ln r + omega0^2 r - 2/r and the phi coefficient is -8 beta ln r / r - 4 beta/r + 3 omega0^2 - 4/r^2: neither is the constant that Eq. 2.162 needs. [calc] Mathematician and contrarian right. With omega = 2(beta - 1/r) (Eq. 2.164) the phi' coefficient is the constant 2 beta, but the phi coefficient is 4 beta/r - 2/r^2, a function of r, not a constant "Hooke's-law" wave number. [calc] Mathematician right; so the failure is that the printed set is not consistent, not that every coefficient is wrong.
4. Electrogravitic size. 9.8 / 8.98755e16 = 1.09e-16 V/m for phi(0) = 1 V. [calc] Physicist and mathematician agree. phi(0) is not given in Felker (called "a fundamental voltage"), so this is an illustration, not a prediction.
5. Photon mass quoted by Felker. 1e-60 kg / 1.78266e-36 kg per eV/c^2 = 5.6e-25 eV, well under the 1e-18 eV bound. [calc] Physicist right; the quoted size is allowed but indistinguishable from zero. (The physicist's text has a garbled "5.6e-25 eV/c^2*... check" phrase; the number is correct.)

Per-claim arbitration

Claim 1. The Einstein field equation is wrong because zero torsion violates a dual Bianchi identity; the connection must be antisymmetric

• Agreement: All three: preface Eqs. 3-7 infer an antisymmetric connection from an antisymmetric commutator label pair, a non sequitur [read]. Torsion is only the antisymmetric part of the connection. If the connection were antisymmetric, the geodesic equation would lose its gravity term (symmetric contraction of an antisymmetric object is zero) and flat space in polar coordinates (Gamma^r_{theta theta} = -r) would be impossible [memory, high; textbook]. The cyclic sum (Eq. 15) is not equivalent to a single contraction (Eq. 16); with T = 0 the true first Bianchi identity holds, whereas Eq. 16 would say Ricci = 0.
• Disagreement: wording only. Physicist says "internally inconsistent"; mathematician adds "refuted by calculation" for "GR violates geometry"; contrarian splits strong/weak form. The mathematician adds a sharper point: for symmetric Gamma the Hodge dual Gamma-tilde built from epsilon times Gamma is identically zero, so the dual identity reads 0 = 0, and Eckardt tests a different object (the ordinary Riemann contraction). I accept this as the strongest argument; it is consistent with the physicist's and contrarian's findings and explains why the tests print "Bianchi identity o.k." [read, per physicist, tables 4.4.1-4.4.2].
• Stronger argument: the mathematician's, because it is a derivation rather than a citation. The contrarian's concession is also kept: torsion is a legitimate extension (Cartan; Einstein-Cartan-Sciama-Kibble) [memory, high]; the torsion-free connection is a choice, not an experimental result at every scale. Evans is right only about that.
• Never tested / tested / dismissed: It is a mathematical claim, so it was neither tested nor dismissed without testing; it was criticised on substance in print (Rodrigues and Souza 2005, Bruhn 2008, Hehl and Obukhov 2008, per [web] Wikipedia; those papers were not read by any reviewer; a related critical note, Giglio and Rodrigues, "One More Assessment of 'ECE Theory'", arXiv:1109.5403, concerns exactly the symmetries of Riemann-Cartan connections and contortion [web] https://arxiv.org/abs/1109.5403, abstract only).
• What would settle it: Compute the actual dual objects for a torsion-free metric (Hodge dual of Gamma is zero) and show the Eq. 16 contraction equals Ricci for all 62 metrics; or ask the author to state the dual identity as a theorem with a coordinate-independent proof.
• Final verdict: Strong form (EFE geometrically invalid; connection must be antisymmetric; Ch.4 shows violation): Internally inconsistent, and the Ch.4 "test" is Refuted by calculation (it is circular). Weak form (spacetime may carry torsion; GR is the zero-torsion case): Plausible, untested, but that is standard Einstein-Cartan theory, not Evans's claim.

Claim 2. 62 metrics all violate the identity; GR, black holes, gravitational waves, Big Bang and dark matter are invalid; Schwarzschild misderived (Crothers)

• Agreement: Ch.4 inherits Claim 1's error. Crothers's alpha argument is about a coordinate choice: Schwarzschild's 1916 R = (r^3 + alpha^3)^(1/3) is a monotonic relabelling; alpha is an integration constant fixed to 2GM/c^2 by the Newtonian limit [memory, high]. "Ric = 0 contains no matter, so it is meaningless" is a misreading (vacuum exterior matched to an interior such as Tolman-Oppenheimer-Volkoff). "The Schwarzschild radius is not a distance" is textbook (areal coordinate). Ch.4 itself reports that some line elements failed the cyclic and metric-compatibility checks (4.12-4.13), which are identities for any Levi-Civita metric, so those failures point to input or code errors [read, per mathematician and contrarian; the individual tables were not read by me or by them].
• Disagreement: (a) the physicist and contrarian label the black-hole/GW impossibility "Tested, not supported" (observation contradicts it); the mathematician labels Crothers's argument "Refuted by calculation" for Secs. 3.2 and 3.4 only. (b) the contrarian found, in the primary source, that Schwarzschild wrote that the solution "contains only the constant alpha that depends on the value of the mass at the origin" [web] https://arxiv.org/abs/physics/9905030 (taken from the contrarian's check; I did not re-open it), contradicting claims.md's wording that alpha is "undetermined and not 2m". (c) the physicist's note that "gravitational waves" could not have been addressed in a 2010 book is correct but secondary.
• Stronger argument: the mathematician's for the algebra (the reviewer who read Ch.3 Secs. 3.2 and 3.4) and the physicist/contrarian's for the observational standing. They do not conflict: calculation disposes of the written arguments; observation (LIGO GW150914 [web] https://en.wikipedia.org/wiki/First_observation_of_gravitational_waves; EHT; pulsar decay [web] https://en.wikipedia.org/wiki/Tests_of_general_relativity) shows the conclusion is also contradicted by data the book never faced. Crothers's gravitational-wave section was not read by anyone, so that part is unassessed.
• Never tested / tested / dismissed: Evans/Crothers ran no experiment; GR's black-hole and wave predictions were independently tested by LIGO/Virgo, EHT, VLBI and pulsar timers. The impossibility claim is therefore "tested, not supported" in the sense that mainstream measurements contradict it, not because anyone designed a test against Crothers. Whether Crothers's papers were rebutted or merely not engaged is unknown from the folder [memory, low; unverified, no verdict depends on it]. The dark matter link is not argued in any passage the reviewers read; dark matter is a genuinely open problem on other evidence [memory, high for the evidence types].
• What would settle it: Compute Ric and the Kretschmann scalar for Schwarzschild in Schwarzschild's R coordinate and in Kruskal coordinates (regular at R = alpha); and re-run Eckardt's Maxima on one metric to confirm that "violation" equals Ricci, flagging which metrics fail 4.12-4.13.
• Final verdict: Ch.4 "62-metric test": Refuted by calculation (circular). Crothers's "black holes and gravitational waves are mathematically impossible": Tested, not supported (observations), with the underlying argument Internally inconsistent (misreads vacuum solutions and Schwarzschild's own text). "Big Bang and dark matter invalidated by this": Not testable as stated (no argument links them).

Claim 3. Electromagnetism is the torsion of spacetime (A = A(0) q, F = A(0) T); ECE unification; Maxwell-Heaviside and gauge theory incomplete

• Agreement: The ansatz is a postulate, not a derivation (F = A(0) T is a definition: any geometric 2-form can be scaled by a constant). The spin connection omega is a free function, "must ultimately be determined experimentally" [read, CEFE p.40]; where omega = 0 it equals Maxwell. No ECE-specific prediction was tested by anyone found. Critics (Hehl and Obukhov, Bruhn on O(3) Lorentz invariance, Rodrigues and Souza on the tetrad postulate) are known only through [web] Wikipedia; nobody read them.
• Disagreement: physicist: "Internally inconsistent" (units, missing right-hand side of the Bianchi identity, Lorentz invariance). Mathematician and contrarian: "Not testable as stated" with inconsistency "plausible but unverified". The physicist's unit objection (dimensionless tetrad versus inverse-length torsion) is partly answered by the mathematician: A(0) in V s/m makes the units work, so I do not count units as an inconsistency. The physicist's Bianchi point (nonzero right-hand side D ^ T = R ^ q) relies on a web summary, not a reading of Evans's derivation. The Lorentz-invariance objection is Bruhn's, via Wikipedia; unverified by me.
• Stronger argument: the mathematician and contrarian, because they restrict themselves to what the text read supports. A claim of internal inconsistency needs a located contradiction, and none was located for Claim 3 beyond the tetrad-has-a-Lorentz-index-but-A-does-not reduction issue, which the mathematician flagged at low-to-medium confidence.
• Never tested / tested / dismissed: Never properly tested. ECE's distinctive predictions (B(3) from the inverse Faraday effect, a gravitational Sagnac effect, polarisation change near a white dwarf) were not tested by anyone found [web] https://en.wikipedia.org/wiki/Einstein%E2%80%93Cartan%E2%80%93Evans_theory; rejection was on mathematical grounds, not failed experiments. The Faraday-disk argument (Sec. 2.7, "no explanation in the standard model") is wrong: the Lorentz force on the moving conductor explains it [memory, high; Griffiths]. This is a bad-argument call against the researcher.
• What would settle it: A single ECE equation set with omega fixed by a stated rule, giving one number (for instance a Sagnac shift or a polarisation rotation) that differs from Maxwell plus GR; failing that, a line-by-line check of the tetrad-to-A reduction against Hehl and Obukhov.
• Final verdict: Strong form (EM is the torsion; Maxwell-Heaviside and gauge theory incomplete): Not testable as stated (free spin connection, postulated ansatz); critics' charge of internal inconsistency is reported but not verified here. Weak form (a geometric origin for EM, as in Weyl and Kaluza-Klein): Historically superseded but instructive.

Claim 4. Spin-connection resonance of the Coulomb/Poisson equation; explains Tesla; Bedini motor draws power from spacetime; electrogravitic E = phi(0) g/c^2

• Agreement: all three: real-r equation has no oscillation (Calc 2); oscillator appears only after the complex "Euler transform" kappa r = exp(i kappa R); no energy budget; a resonance amplifies a response and does not create energy; Tesla coils resonate through LC circuit physics without omega [memory, high]. Electrogravitic is dimensionally V/m and about 1e-16 V/m for phi(0) = 1 V (Calc 4).
• Disagreement: the physicist did not check Eqs. 2.161-2.165; the mathematician and contrarian did and found they do not follow (Calc 3). The physicist says the free-power claim is "Refuted by calculation"; the mathematician says "Never properly tested", and refuted only in the sense that no energy budget is supplied; the contrarian says both. Note that the physicist treats Evans's statement that resonances "cannot be explained using the standard model" as false; the contrarian agrees on the Faraday disk analogue. Evans's text says "No violation of conservation of energy and momentum occurs" [read, p.41] but gives no accounting.
• Resolution on the energy claim: two separate things must not be merged. (1) The mathematical derivation contains no energy source term and a static curl-free field does no cyclic work; accepted laws applied correctly therefore do not allow it to power anything: this is a calculation-level conclusion about the derivation. (2) Whether a particular device (Bedini motor, SCR circuits) shows COP > 1 is a measurement; no neutral controlled test is on record in the folder or in the Wikipedia summary [web] https://en.wikipedia.org/wiki/Einstein%E2%80%93Cartan%E2%80%93Evans_theory ("no documented replication"). I attempted a web check for independent Bedini tests; the search budget was exhausted, so that check failed and the statement "no neutral test on record" remains limited to the sources named. The device claim was dismissed on theoretical grounds without a test, but the claimants also supplied no input/output data in these books (papers 63, 94 unread).
• Stronger argument: the mathematician's, who verified the algebra symbolically and noted that Eq. 2.162 does not follow from 2.160-2.161; this also makes the contrarian's "tested/never tested" distinction hold.
• What would settle it: Calorimetric or calibrated power-meter input/output measurement of the Bedini circuit, run by someone who does not sell it, with the full-cycle energy accounted (including battery state of charge); on the theory side, solve Eq. 2.156 for real r with a stated physical charge distribution and show a resonant solution at real radius.
• Final verdict: Resonance derivation: Internally inconsistent (Eqs. 2.161-2.165 do not produce 2.162; the "resonance" lives in complex r). Free power from spacetime: derivation Refuted by calculation (no source term; energy conservation), device claim Never properly tested (dismissed without a test; no neutral test offered). Electrogravitic E = phi(0) g/c^2: Not testable as stated (phi(0) unspecified; about 1e-16 V/m for 1 V).

Claim 5. Uncertainty principle refuted and replaced by (box + kT) q = 0; photon has B(3) longitudinal field and mass; electroweak masses without Higgs

• Agreement: the preparation form sigma_x sigma_p >= hbar/2 is a Fourier/Cauchy-Schwarz theorem [memory, high]; Felker's replacement x ^ p -> hbar is a wedge-product relation, a different kind of statement; the "Lemma" box q = R q followed by R = -kT is a definition and a postulate, and Klein-Gordon is obtained only by setting R = (mc/hbar)^2 by hand [read, Felker]; q is a tetrad, not a complex normalised wavefunction. The Croca refutation is quoted from Evans in Felker and unconfirmed; B(3) from the inverse Faraday effect, a real effect explained in mainstream theory without a longitudinal photon [memory, high].
• Disagreement: (a) uncertainty: mathematician "Refuted by calculation" plus "Tested, not supported"; physicist and contrarian "Tested, not supported". I take the physicist/contrarian position: the claim as stated is experimental ("refuted by the Croca group"), and the experimental record contradicts it. The mathematical theorem is the reason the claim cannot be a refutation, and the Rozema 2012 results concern error-disturbance versus the preparation relation (Ozawa correction), a different statement [web] https://arxiv.org/pdf/1208.0034 per the mathematician's summary read; not re-opened. (b) wave equation: physicist and mathematician "Internally inconsistent"; contrarian "Not testable / never tested". (c) no Higgs: physicist "Tested, not supported"; contrarian "Never properly tested". The Higgs discovery (2012, LHC) [memory, high] is tension, not a test of an Evans derivation, since the book gives no numerical mass derivation; no reviewer read Felker Ch.11-12 in detail.
• Stronger argument: on (b) the physicist and mathematician: the defect is located (R is defined, then equated to -kT, then set to a mass by hand), so "inconsistent as a derivation" is accurate; "not derived" is the safer phrasing. On (c) the contrarian, who keeps tested and untested apart.
• Never tested / tested / dismissed: Uncertainty: tested enormously and supported. Croca's claim: unreplicated; I tried to verify Croca's status on the web and the search budget was exhausted, so it stays [memory, low] ("not accepted") and no verdict depends on it. Evans wave equation and masses: no independent test and no number derived.
• What would settle it: Show a Gaussian wavepacket with sigma_x sigma_p below hbar/2 in any measured data (impossible for a normalised wavefunction); or derive from (box + kT) q = 0 a single numerical prediction (the W/Z mass ratio, or electron g-2) and compare with measurement. On the photon mass: a measured nonzero value would need a different experiment; current bound is below 1e-18 eV.
• Final verdict: "Uncertainty principle is refuted": Tested, not supported. "Evans wave equation derives Dirac, Klein-Gordon, Schroedinger": Internally inconsistent as a derivation (postulated R, hand-set mass, tetrad treated as wavefunction), confidence medium (Felker only sampled). B(3) longitudinal photon field: Not testable as stated. Photon mass tiny and nonzero: Plausible, untested (bound below 1e-18 eV [web] https://pdg.lbl.gov/2026/listings/rpp2026-list-photon.pdf via reviewers' search snippets; my own fetch of the 2025 PDF returned no readable text, so this check is only partly confirmed; a mass of about 5.6e-25 eV, per Felker's 1e-60 kg, is allowed but not distinguishable from zero). Electroweak masses without a Higgs: Never properly tested (no mass derived; in tension with the 2012 Higgs discovery).

Claim 6. Academic physics ignored or abused these results ("belief retention"); readership shows acceptance

• Agreement: all three: not a physics claim; web hit counts are not replication; "crude personal abuse" at Physica B (paper 93) is Evans's own statement with no referee report in the folder; and "ignored" is contradicted by published technical critiques and an editorial (t Hooft 2008), per [web] https://en.wikipedia.org/wiki/Einstein%E2%80%93Cartan%E2%80%93Evans_theory, where ECE is recorded as in Foundations of Physics Letters 2003-2005 and then closed by the editor with ECE absent from standard archives. The Physica B story is not mentioned there.
• Disagreement: none on the verdict. The contrarian alone separates the two sub-histories: geometry claims were criticised on substance (not ignored); device claims were dismissed without a test but no test was ever offered by the claimants.
• Stronger argument: the contrarian's split; it keeps the "dismissed without testing" category visible without implying suppression of a well-posed claim. Equal rigour for the researcher: self-description as "The British and Commonwealth Civil List Scientist" is a title, not evidence; "fully accepted" and "abuse" are used together unfalsifiably [read].
• What would settle it: the referee reports for paper 93 and the Foundations of Physics editorial correspondence; a download-count versus citation-by-independent-groups comparison.
• Final verdict: Not testable as stated (sociological). Where specific: "ignored without engagement" is contradicted by the published critiques (so false for the geometry papers), "abused at Physica B" is unverified, and the device claims were dismissed without testing while never being offered for neutral testing.

QE revision (after qe.md, verdict MINOR)

Page convention for this section: PDF page numbers (CEFE printed page = PDF minus 8; Felker printed = PDF minus 6). Numbering: this file keeps its original claim numbers 1-6 and adds Claim 7. claims.md was renumbered after QE; mapping: arbiter 1 = claims 1; arbiter 2 = claims 2 and 3 (Ch.4 test; black holes/waves/dark matter); arbiter 3 = claims 4; arbiter 4 = claims 5; arbiter 5 = claims 6; arbiter 7 = claims 7. Arbiter 6 (sociological) is the unnumbered context item in claims.md.

Sources for this section: qe.md; deep/book-CEFE-print-070c33.md (Crothers Secs 3.5-3.11, PDF pp.110-136; Ch.4 Secs 4.1-4.3, PDF pp.141-152); deep/book-Evans-Equations-Rev3-b8855b.md (Felker Chs 9-12, PDF pp.203-250). My own [calc] checks were run in .venv. I did no new web searches; web items below are those the deep reads recorded, and I did not re-open them.

New evidence

1. LIGO and Gravity Probe B (CEFE PDF pp.130, 134). Crothers wrote: "The LIGO project ... have not detected gravitational waves [74]. They are destined to detect nothing" and that Gravity Probe B did not detect frame dragging [read: pp.130, 134]. This is a dated prediction that was tested afterwards: GW150914 (14 Sep 2015, announced 11 Feb 2016) [web: https://en.wikipedia.org/wiki/First_observation_of_gravitational_waves]. His source [74] is an initial-LIGO continuous-wave search that reported only upper limits (arXiv 0710.4898 [web, per deep read]); his source [75] is an advocacy web page [read: p.138]. Gravity Probe B's final result (Phys. Rev. Lett., May 2011): frame dragging measured -37.2 +/- 7.2 against predicted -39.2 milliarcseconds/year, about 19% precision [web: https://en.wikipedia.org/wiki/Gravity_Probe_B, per deep read]. The "not detected" statement was fair as a 2008-2010 status report and is superseded; it was never a failed test of GR. Correction to physicist.md item 5: the 2010 book did address LIGO; it predicted a null result. (I note the physicist's file was left unedited, since Step 2 is not redone under MINOR.)
2. Crothers Secs 3.5-3.9 now assessed [read: deep read]. Infinite density forbidden by special relativity (3.5): the algebra is right (density factor 1/(1-v^2/c^2)) but only closes one route to infinite density; "GR cannot violate SR" holds locally only. Escape velocity (3.4, 3.6): a fair criticism of popular wording, not of GR, whose horizon is causal. No two-body solution (3.7): the premise is false; Crothers himself quotes the exact Majumdar-Papapetrou multi-body solution (p.115) [read], and an existence theorem (Choquet-Bruhat 1952), numerical relativity and Hulse-Taylor decay exist [memory, high]. Energy localisation (3.8): his replacement Eq. 3.26, G/kappa + T = 0, is Einstein's equation rearranged, and with T = 0 it gives G = 0, hence R = 0 and Ricci = 0, the vacuum equation he says is impossible [calc]. Kretschmann scalar, Doughty acceleration, "signature change" (3.9): K = 12 alpha^2/r^6 is the textbook value [calc]; his Doughty acceleration is twice the value his own formula gives [calc]; (+,-,-,-) and (-,+,-,-) have the same signature, so nothing changes sign in the signature [calc]; Hagihara 1931 implies geodesic incompleteness, which supports the Kruskal extension. His Weyl 1944 "no such tensor" claim states no theorem and could not be checked [memory, low]. The Max Planck "admission" is private correspondence posted on his own site [read: p.138].
3. Ch.4 definitions and the "failed" tests (CEFE PDF pp.141-152). Every "charge density" and "current density" table entry is a component of minus the contravariant Ricci tensor, R^kappa_mu^{mu nu} = -R^{kappa nu} for the Christoffel connection (rho = R^00; class 1 = diagonal; classes 2 and 3 = off-diagonal pairs, the same three numbers reordered; no R^{0i}) [read; calc, confirmed on three of the book's own tables, and QE's Kasner table]. The three "current classes" carry one identical label in the book. Correction to my own earlier text: I wrote that Ch.4 "reports that some line elements failed" the cyclic and metric-compatibility checks. The text says only "several well known line elements failed" (p.144) and names none; all 62 tables print "o.k." for both tests [read: pp.151-503, per deep read]. Both tests are identities for any Christoffel connection of a symmetric metric, so a true failure could only come from input or code error. The open question "which of the 62 failed" is therefore answered: none of those printed. Two concrete input errors are now documented: the book's "Kerr" metric is the first-order Lense-Thirring form with a probably doubled cross term (not exact Kerr, which is Ricci-flat), and QE's Kasner case (see 4) and the Kruskal table (nonzero Ricci for a vacuum metric). The book's "vacuum solutions by definition assume zero curvature" (p.142) is wrong: its own Schwarzschild table prints a nonzero Riemann component with Ricci zero (p.155) [read].
4. Kasner and FLRW tables (QE Check 3, qe.md). The book prints J1 = -p1(p1+p2+p3-1) t^(-2p1-2), which vanishes exactly when the Kasner vacuum conditions hold, and for FLRW its "charge density" is -3a''/a, the Ricci component I computed in Calc 1 [read, per QE]. The book says the Kasner conditions can "probably" be met only trivially; false, since (-1/3, 2/3, 2/3) has sum 1 and sum of squares 1 [calc]. It then plots p = (1,-1,0), which has sum 0 and is not a vacuum Kasner solution [read, per QE]. A Ricci-based "violation" for an invalid input is just what the circular test would give.
5. CEFE Sec. 2.8, g-factor and Lamb shift (PDF pp.54-61; QE Check 2). The book gives g(ECE) = 2.002323(49) and g(exp) = 2.0023193043718(75), admits a difference of 0.000004, and calls QED precision tests "of no utility and no meaning" [read: PDF pp.55-56, per QE]. Its Eq. 2.223 gives g = 2 + alpha/pi + alpha^2/(8 pi^2). With alpha = 7.2973525693e-3 [calc]: g = 2.0023234939, which is 4.19e-6 above experiment, 5.6e5 times the quoted experimental uncertainty (7.5e-12). The book's own alpha error bar (0.007297(34), 0.5% relative) is about 10^4 times larger than alpha's real uncertainty (about 1.8e-7 relative, per QE; I did not re-derive that). Since 2019, h is exact in the SI [memory, high], so "limited by the accuracy of h" no longer applies at all. The Lamb-shift figure 0.0353 cm^-1 uses a vacuum radius r that the book states (r = 1.69e-7 m, Eq. 2.243) but never justifies (corrected in Round 2 below; I earlier wrote "unstated"); the measured 1057.8 MHz is 0.03528 cm^-1, so the agreement is a choice of r, not a prediction [calc, per QE]. The Schwinger term alpha/pi is already standard QED; the extra alpha^2/(8 pi^2) term has no relation to the actual second-order QED coefficient (-0.328 alpha^2/pi^2 [memory, medium]), though I did not check that.
6. Felker Chs 9-12 (PDF pp.203-250). - Boson masses. The text says only: "The results of the calculations are the experimentally observed masses of 78.6 GeV/c2 (W) and 89.3 GeV/c2 (Z)" [read: p.245, page image]. No calculation is shown; W(0) is never given a value; hbar and c alone cannot give a mass (hbar/c has units kg m). The framework itself says the mass is an adjustable input to the curvature: "Mass m is adjustable ... R the curvature is then fixed for any given mass" [read: p.240]. The quoted values differ from measurement (about 80.4 and 91.19 GeV [memory, high]) by 2.2% and 2.1% [calc], and they are presented as "observed", which is not the case. - Minimum volume mV0 = k hbar^2/c^2 [read: pp.203, 214-219]. It follows from the assumed identification kT = (mc/hbar)^2 with T = m/V0, using the nonstandard k = 8 pi G/c^2 with T a mass density [read: p.214]. V0 is thereby defined, so "m V0 is a universal constant" is true by definition. The printed electron figure V0 = 1e-79 m^3 does not follow: I get 2.54e-81 m^3 [calc], 40 times smaller; the "radius 4e-13 m" is the reduced Compton wavelength, unrelated to V0; the disk thickness 2.18e-56 m is back-fitted; the sphere arithmetic drops a cube (p.217). Felker labels the shape speculation as his own [read: p.216 n.46]. - Microscope h/25 [read: pp.220-221]: Delta x = lambda/50, Delta p = 2h/lambda give h/25. The book quotes Heisenberg's bound as h/2 (p.221) but as hbar/2 on p.222. The correct bound hbar/2 = h/(4 pi) = 0.0796 h, so h/25 = 0.04 h is only about half of it, and the quantity is a resolution times a one-photon kick, not the spread of one prepared state. No apparatus or data is given. Afshar is cited with a URL and no experimental details, and "Shariar" for Shahriar; its interpretation is disputed [memory, medium; not web-checked]. Croca is cited only through Evans's email [read: p.225 n.51]. - B(3) evidence. A single footnote: "experimental confirmation ... in the inverse Faraday effect of an order of a million gauss with pulsed circularly polarized lasers in ultra-dense plasma ... no doubt that the B(3) field exists", citing Evans's own papers and Adv. Chem. Phys. vol. 119 [read: pp.228-229]. No laser parameters, no authors of the experiment. The inverse Faraday effect is real (first seen in the 1960s [read: p.232]) and has a conventional description; the book gives no argument that it needs B(3). Megagauss laser-plasma fields are reported in mainstream plasma physics [memory, low; not checked]. - Smaller points: the "Principle of Least Curvature" R0 = -(mc/hbar)^2 is a definition and E = hbar c sqrt|R0| restates E = mc^2 [read: pp.210-211]; the book's neutrino-oscillation and muon-decay sections give no number [read: pp.247-250]; "The weak force holds some particles together" is wrong [read: p.239; memory, high]. Felker says the speculative parts are his, and QE notes that he states Evans "did not correct" the book [read per QE, pp.20-21].
7. Other QE items I accept, not individually re-verified: CEFE Ch.5 (Pendergast, PDF pp.526-533) accepts light bending, lensing and the Sgr A* orbits, so the book is not uniform in rejecting standard observations [read, per QE digest]; its Sagnac derivation recovers the standard result; the claim that the Eddington-precision criticism is in Ch.1 not Ch.5 is a facts.json correction.

Changed and confirmed verdicts

Answers to the open questions raised earlier

• "Which of the 62 metrics failed the cyclic/compatibility checks?" None of the 62 printed tables; the claim is unsupported by the book's own output.
• "What does Crothers's gravitational-wave section argue?" See New evidence 2; each argument has a textbook answer, listed there.
• Still open: papers 63, 93, 94 and the critics (Hehl and Obukhov, Bruhn, Rodrigues and Souza) are not in the folder; Croca and Bedini independent-test status are unverified (search budget was exhausted earlier); the Weyl 1944 theorem content is [memory, low]; Eckardt's torsion simulations for galaxy curves are not in the folder.

What would settle the new items

• Claim 7: a derivation of the vacuum radius r in the Lamb-shift formula (the book states r = 1.69e-7 m but gives no derivation), and the g-2 coefficient at order alpha^2 compared with the measured value; one line of arithmetic already puts g(ECE) 4.2e-6 away.
• Boson masses: the actual values of W(0) (or the W field components) that enter Eq. (10); with them, one multiplication would show whether 78.6 follows.

Round 2 (deep read of CEFE Ch.2 Secs 2.7-2.9, PDF pp.49-66)

Source: deep/book-CEFE-print-070c33.md, section "QE round 2" (printed page = PDF page minus 8). I did not re-run the calculations; I rely on that deep read's [calc] and [web] tags (NIST CODATA pages: https://physics.nist.gov/cgi-bin/cuu/Value?alph, ...?gem, ...?h), which I did not re-open. Checked by me against the book's printed formula: Eq. 2.223 expands as 2 + alpha/pi + alpha^2/(8 pi^2), matching my earlier Calc.

Result: no verdict category changes. The deep read confirms every verdict above and sharpens the reasons. Corrections and additions:

1. Correction to my own text (and QE/primer): the Lamb-shift radius is stated. Eq. 2.243 prints "r = 1.69 x 10^-7 m" [read: PDF p.59]. It has no derivation and is not called a fit. It equals about 3190 Bohr radii, far outside a 2s orbit (mean radius about 3.2e-10 m); with r = 4a the same formula gives 28 cm^-1 and with r = a 113 cm^-1, hundreds to thousands of times the measured 0.0353 cm^-1 [calc, per deep read]. The value that exactly matches 1057.8 MHz is 1.694e-7 m. So r works as a free fit parameter and the Lamb-shift agreement is not a prediction. Verdict for Claim 7 stays Tested, not supported. The Lamb-shift "angular dependence" is dropped (cos(theta) = 1) and delta E proportional to 1/r at fixed r has none of the real n and l dependence [memory, high, per deep read].
2. The ECE g-factor is worse than Schwinger's one-term value. g(Schwinger) = 2 + alpha/pi is 3.5e-6 above experiment; ECE's extra +alpha^2/(8 pi^2) = +6.7e-7 has the wrong sign against true second-order QED (about -3.5e-6 in g) and takes the gap to 4.19e-6. Standard QED through alpha^2, alpha^3, alpha^4 lands at -3.0e-8, +1e-10, -1e-11 from experiment [calc, per deep read; QED coefficients memory, high for the first three]. This replaces my earlier hedged remark that I had not checked the alpha^2 term. The book's printed QED series (Eq. 2.227) uses (alpha/2 pi)^2 where (alpha/pi)^2 is standard [memory, high].
3. The "inconsistency in NIST" is refuted by calculation. Using the book's own inputs, alpha = 7.2973525546e-3, within 5e-12 (0.2 of NIST's quoted 2.4e-11 uncertainty) of the NIST value the book quotes; the "mismatch" comes from truncating alpha to 0.007297 in Eq. 2.213 [calc, per deep read]. The error bars "(34)" on alpha and "(49)" on g are not reproducible: (34) is about 2e4 times larger than the book's stated rule (relative uncertainty of h) gives, and none of three propagations gives 4.9e-5 [calc]. With the book's own error bar the 4e-6 gap is "within error" only because that bar is 6.5e6 times the experimental one. This updates my "about 10^4" figure to about 2e4. Smaller slips in the same section: Eq. 2.207 labelled hbar is h; 2.229 R_K equalities wrong; 2.249 is 3.47e-14 m not 3.48e-13 m (the book's own "ten times the classical radius" fits the correct figure); 2.231 versus 2.240 (1/4pi versus 1/(2 pi^(3/2))) and 2.253 versus 2.242 (factor 2) disagree [calc, per deep read]. Alpha "limited by h" no longer holds since the 2019 SI redefinition [memory, medium; deep read]. Claims that alpha "drops to 1/128" at high energy are backwards (alpha rises) [memory, high].
4. Faraday disk (Sec. 2.7): the book's "no explanation in the standard model" is false (the Lorentz force gives emf = B Omega R^2/2 [memory, high]); the book supplies no emf formula or number, and its Eq. 2.193 applied to the A(2) component gives zero, only A(1) gives Eq. 2.194, so the printed step is inconsistent with the result [calc, per deep read]. This supports my existing bad-argument call under Claim 3; no category change.
5. Sagnac (Sec. 2.7): a restatement, not a derivation. The standard delay 4 Omega A/c^2 is the input; omega1 = c/r is solved to return it; no new prediction (no second-order term) [calc, per deep read]. The book says its gravity-optics effects are "not observable" in the laboratory (p.54), a concession relevant to Claim 3's "no testable ECE number". Claim 3 stays Not testable as stated.
6. Sec. 2.9 list of 36 "advances" (PDF pp.61-66): all 36 are present and none has a number, equation or citation [read]. Several quantitative-sounding items (galaxy rotation curve, "incorrect by orders of magnitude") rest on nothing printed on these pages. The statements that the Higgs "has not been verified" date the text to about 2008 (it places itself "five years since inception in Spring 2003"); the Higgs-like boson was announced in July 2012 [memory, high]. This is a dating note, not a failed prediction by the book, and does not change Claim 5's treatment (the W/Z numbers come from Felker, not from here).

Verdict table changes (Round 2): none. Claim 7 stays Tested, not supported, with the NIST-alpha sub-claim now stated as Refuted by calculation and the error bars as Internally inconsistent with the stated method. Claims 3 and 4 (Faraday disk, Sagnac sub-arguments) stay as before, with the added reasons above. Wording changes were made in the verdict table row 7, Claim 7 item in the Corrected claim list, the "What would settle" line and the in-place Lamb-shift wording, all to read "stated but unjustified" instead of "unstated".

Corrected claim list

1. EFE with torsion-free connection is not shown to violate any dual Bianchi identity: the preface's antisymmetric-connection step is a non sequitur, and Eq. 4.1 with T = 0 is the vacuum equation Ric = 0. Torsion is a legitimate generalisation (Einstein-Cartan), not a refutation of GR.
2. The Ch.4 computer-algebra "violations" equal the Ricci tensor (confirmed from the book's definitions and its RN, "Kerr", Kasner and FLRW tables), so they restate the Einstein equation with matter. The book says several line elements failed metric compatibility and the cyclic identity but names none; all 62 printed tables say "o.k." (QE revision). Crothers's arguments do not establish that black holes, gravitational waves, the Big Bang or dark matter are impossible: the book's own 2010 prediction that LIGO is "destined to detect nothing" was falsified by GW150914 (2015); his replacement field equation reduces to Ricci = 0 when T = 0; his Doughty acceleration is off by a factor 2 and his "signature change" is not one. Gravity Probe B's "not detected" was true as a 2008-2010 status and was superseded in 2011. Dark matter from torsion rests on simulations not in the folder. Input errors in Ch.4 are documented (the "Kerr" metric, Kruskal, Kasner at p = (1,-1,0)).
3. ECE postulates F = A(0) T with a freely chosen spin connection; where omega = 0 it is Maxwell, elsewhere it makes no number that differs from mainstream physics. It has not been tested. Its Faraday-disk and Sagnac sections give no new number (Sagnac recovers the standard 4 Omega A/c^2 by solving for omega1; the disk emf is never computed), the book says its gravity-optics effects are unobservable in a laboratory, and the "no standard explanation" claims are false.
4. Resonance of the Poisson equation with a spin connection appears only after continuing the radius into the complex plane; the general-case equations as printed do not follow from each other; no energy budget is given. Device claims (Bedini) have no neutral test on record. E = phi(0) g/c^2 would be of order 1e-16 V/m for phi(0) = 1 V.
5. The uncertainty relation is a theorem and experimentally confirmed; the microscope "h/25" argument misquotes the bound and measures resolution, not state spread. The Evans wave equation postulates R = -kT and sets the mass by hand; the minimum-volume relation m V0 = k hbar^2/c^2 is a definition and its electron numbers do not follow from it (2.5e-81 m^3 from the formula, not 1e-79). Photon mass is bounded below 1e-18 eV. Felker states, with no derivation shown, W and Z masses of 78.6 and 89.3 GeV, called "observed" but about 2% below the measured 80.4 and 91.19 GeV; the Higgs boson was found in 2012. The B(3) "evidence" is one footnote about the inverse Faraday effect.
6. The record shows that the geometry work received specific published criticism, the device claims received no test either way, and the Physica B "abuse" story is unverified.
7. The book's own Eq. 2.223 gives an electron g-factor of 2.0023235, 4.2e-6 above the measured 2.0023193044, about 5.6e5 times the experimental uncertainty; its quoted error on alpha is about 2e4 times too large (Round 2; QE had about 1e4); its Lamb-shift figure depends on a radius (1.69e-7 m, about 3190 Bohr radii) that is stated but not derived. The ECE alpha^2/(8 pi^2) term makes g worse than Schwinger's alpha/pi alone, and the claimed inconsistency of NIST's alpha is a rounding artefact (the book's own inputs reproduce it to 0.2 sigma). The claim that ECE reproduces g-2 and the Lamb shift without QED, and that QED precision is meaningless, is not supported.

What would settle it (summary)

• Claim 1: evaluate the true Hodge-dual connection for a symmetric Gamma (it is zero) and the Eq. 16 contraction for Schwarzschild (it equals Ricci).
• Claim 2: tabulate the 62 metrics against Ric and the two failing identities; Kretschmann scalar of Schwarzschild across R = alpha in Kruskal coordinates.
• Claim 3: one ECE-specific number (Sagnac, polarisation, white-dwarf lensing) with omega fixed in advance.
• Claim 4: neutral input/output calorimetry of the Bedini circuit; real-r resonant solution of Eq. 2.156.
• Claim 5: any dataset with sigma_x sigma_p < hbar/2; a derived W/Z mass ratio from the wave equation.
• Claim 6: the paper 93 referee reports.

Open questions

• What exactly did Hehl and Obukhov, Bruhn and Rodrigues and Souza prove, and does any of it address the CEFE "Hodge-dual identity" of 2007-09? (No reviewer read them.)
• Which of the 62 metrics failed the cyclic and metric-compatibility checks in Ch.4, and why?
• What does Crothers's gravitational-wave section (Ch.3) actually argue?
• Do Evans's papers 63, 90 and 94 contain a corrected version of Eqs. 2.161-2.165?
• Has any neutral party measured the Bedini device or the SCR circuits (papers UFT364/382/383 named on Wikipedia)?
• Is Croca's nonlinear-quantum-mechanics result a legitimate anomaly or a misreading of the preparation relation? (Unverified.)
• Genuine open physics near this project (not credit to Evans): the dark-matter evidence versus modified gravity, torsion bounds beyond spin-polarised tests, and the status of strong-field GR beyond the first LIGO and EHT results.

Web verification log (budget 8; search budget exhausted)

• [web] https://arxiv.org/abs/1109.5403 (opened): Giglio and Rodrigues, "One More Assessment of 'ECE Theory'", about connection and contortion symmetry; supports that critics addressed Evans's symmetry confusion; abstract only. Replaces the Step 1 [memory] low item about Hehl/Obukhov and the dual identity: that still stays unverified specifically for the dual identity.
• [web] https://pdg.lbl.gov/2025/listings/rpp2025-list-photon.pdf: fetch returned binary data without readable text; the 1e-18 eV bound stays supported by earlier reviewer snippets from the 2026 listing only. Partly confirmed.
• Croca replication status and Bedini independent test: web search refused (session search limit reached); remain [memory] medium/low and are phrased as unverified above. No verdict depends on them as a finding of fact.
• Other [web] items (Wikipedia pages, Schwarzschild translation arXiv physics/9905030, arXiv 1208.0034) were taken from the reviewers' checks, not re-opened by me.
• Remaining [memory, high] facts (Higgs 2012, Levi-Civita connection, first Bianchi identity, Tesla coil LC resonance, Fourier origin of uncertainty) were not re-checked; no verdict depends on a medium- or low-confidence fact other than those noted.

Verdict table