Gardi-Lori: quality engineering (Step 4b)

QE run 2026-10-03. Inputs read: _text/manifest.json, all 19 digest files in digest/, facts.json, claims.md, physicist.md, mathematician.md, contrarian.md, arbiter.md, primer.md, example.py. Spot checks used the extracted text in _text/ and page images I rendered to scratch space from the read-only source folder (nothing was written there). No deep/ folder exists: this project was reviewed before digests and deep reads were part of the process.

Verdict: MINOR (borderline MAJOR). The review is careful on claims 1-5 and 7. It has one serious factual error: all three reviewers, the arbiter and the primer say the particle paper computes no particle masses. It does. Table 5.1 of "The Geometry of Particles" gives a one-parameter lepton mass formula and compares it with Particle Data Group (PDG) values. That changes the claim 6 verdict from "Never properly tested" to a negative verdict. The change is confined to one claim, I have already reproduced the numbers, and it can be fixed through a "QE revision" in the arbiter and an updated primer. So I grade it MINOR. Escalate to MAJOR if the Spacetime or Electric Universe deep reads (items 2 and 3 below) find further non-identity numerical predictions that the reviews scored as circular identities.

---

1. Coverage

The manifest's 19 chunks are 13 single-chunk documents plus 3 two-chunk documents (PaperColour, Spacetime v2.1, Spacetime v2.3). Coverage is complete. No document was skipped, so no skip reasons need checking.

Minor: the MyART digest file is headed with the PDF file name, but its content is the Dinu-Gardi "electrigen" paper. facts.json records this correctly.

2. Shortcuts

Step 1 ran before digests existed and sampled the large documents. facts.json is honest about this: its how_read fields say "sampled", "first half", "keyword-targeted". The shortcuts are judged below.

3. Missed material (from my own reading of the digests, checked in the text)

1. The lepton mass formula (the most important omission). Geometry of Particles, Eq. 4.10 [read: page image PDF p.14]:

E = E_ν · K⁵ · Σ_{n=1..K} [n³ − (n−1)³] / n⁴

Here E_ν = 2.38925 eV is the energy of the "fundamental" neutrino, and K is the number of wave centres. Table 5.1 [read: text pp.16-17] compares the formula with PDG values. I reproduced every row [calc, this QE]:

What this means: - (a) The model makes testable numbers, so "Never properly tested (no quantitative output)" is wrong. - (b) It has one fitted energy and a free integer K chosen per particle "by best fit" (p.13, steps 1-2). With E roughly proportional to K⁵, neighbouring K values differ by about 19 % near K = 28 (E(29)/E(28) = 1.194) [calc]. So any mass lies within about 10 % of some K, and agreement to about 10 % is guaranteed. The fit has no predictive power. - (c) K = 29 gives 113.3 MeV, which is 7.2 % from the muon mass and closer than K = 28 [calc]. K = 28 was kept because it is a nuclear "magic number". The magic-number match the reviewers called numerology is therefore partly built in. - (d) The neutrino energy of 2.39 eV, and neutrino masses of 0.16 MeV and 17 MeV, conflict with direct and cosmological limits. KATRIN gives m_ν < 0.45 eV (2025) [memory, medium-high; needs web check]. Oscillation data plus cosmology put the sum of the three masses below roughly 0.1-0.2 eV [memory, medium].

Suggested revised verdict for claim 6 (particle part): Tested, not supported. The neutrino-scale masses are contradicted by direct measurement. The charged-lepton fits are post-hoc and carry no evidential weight.

2. Composite proton and neutron, and solar-neutrino-triggered decay (Geometry of Particles pp.18-21 [read]). The proton is described as four electrons plus one positron, and the neutron as five electrons plus one positron. - Naive charge counting gives −3e and −4e, not +e and 0. The paper says only that the parts' charges "cease to be recognized". That is an internal-consistency problem none of the reviews raised. - The paper says beta decay is triggered by solar antineutrinos. The Sun emits electron neutrinos, not antineutrinos [memory, high]. The free-neutron lifetime (about 880 s) is measured in laboratory bottles and beams with no dependence on solar flux [memory, high]. - Electrons inside nuclei were abandoned in the early 1930s because of the nitrogen-14 spin statistics and nuclear magnetic moments [memory, high]. That makes a good "historically superseded" teaching point for the primer.

3. "Validated from hydrogen to calcium" (Spacetime v2.3 text line 1194, p.26-27 [read]). The Rydberg/frequency equation with an interference factor D is said to reproduce photon energies from hydrogen to calcium "in a separate paper [15]". That paper is not in the folder. claims.md lists the companion works not read; this one (Yee's atomic orbitals paper) should be added.

4. Dropping Maxwell's displacement current (Planck V2 Appendix B, p.10; V3 Appendix B, pp.9-10 [read: digest, text]). The paper concludes ∇×H = J (V3), or that displacement current and current density are "equivalent" (V2). Without the ∂D/∂t term there are no electromagnetic waves in vacuum, and charge conservation fails (∇·(∇×H) = 0 forces ∇·J = 0 always) [memory, high]. It is cited to Robert Distinti. It is a striking, checkable error and appears in no review file. It belongs under claim 4 or claim 7.

5. Electric Universe second half: strong force 42,127 N at r_p = 8.663e-16 m, said to be "~137 times" the electric force (digest p.7-8); an inverse-cube "orbital force" keeping the electron at the Bohr radius; h "derived" (pp.10-12). Spot arithmetic: k e²/r_p² = 307 N, and × 137.036 = 42,100 N [calc], so this is the Coulomb force times 1/α by construction. It should be noted under claim 3.

6. CNPS 2018 gravity "three ways" (p.7): G m_e²/r_e = m_e c² α_Ge = 1.97e-56 J. Here α_Ge is the electron's gravitational coupling, which is defined as G m_e²/(ħc). So this is an identity, but it is not mentioned anywhere. Minor.

7. Smaller items not carried forward: - Simhony's NaCl work is cited (in facts and primer; fine). - Juliana Mortenson's "hidden variables in light energy" acknowledgement (V3). - Bryan Steele (credited with the MUA laser example). - The 1960 adoption of "hertz" as the historical origin of MUA (V3 p.5). - MIT 500 pK quoted by Gardi herself as "one order of magnitude" above her floor (V2 p.6). The mathematician has this; the primer does not. - PaperColour's comparisons with S-star orbits and Fermilab bubble-chamber tracks (p.12). - Pear-shaped Ra-224/Ba-144 nuclei offered as support. The physicist covers this.

4. Accuracy spot checks

Additional small accuracy issues: - contrarian.md, claim 4: "one cycle carries about 4e-16 times less than one photon energy". The ratio h/(hf) = 1/f = 2e-15 at 5e14 Hz, so h is about 5e14 times smaller than hf. This does not affect the verdict. - Page-image page numbers for the Spacetime paper do not match across files. claims.md says "pages 7, 8, 28 and 29"; the arbiter says "pages 5-9 and 27-28". Files should state whether they mean PDF or printed pages: the PDF numbers run one ahead of the printed ones.

5. Process

• Source tags. Present throughout and generally honest. The arbiter and primer are explicit when they rely on another reviewer or on Step 1. The source-gap note is repeated in the arbiter and the primer, as required.
• Web checks. These are good for the verdict-bearing facts: the α and a_e CODATA values, Fermi GBM, INTEGRAL, GW170817, EHT M87*, and the PRL on 38 pK. Two verdict-relevant [memory] items remain:
• (i) later binary-black-hole events lacking gamma counterparts. The arbiter tried, failed, and correctly kept it out of the verdict.
• (ii) the KATRIN neutrino-mass bound: physicist [memory, medium, not web-checked], mathematician [memory, medium, unverified]. With the revised claim 6 this bound becomes verdict-bearing, so it must be web-checked.
• Verdict scale. Applied consistently, and the arbiter's conventions section, separating circular identities from refuted physics, is a sound addition. Claim 6 is the exception: "Never properly tested" rested on a false premise.
• Primer. Follows the STEP 4 order (sections 1-9 plus answers). It defines symbols and explains Planck units and α from scratch, so it fits the reader profile. The worked example (sound speed, from salt to the vacuum) is real, step-by-step and backed by example.py. Defects: the claim 6 row of the evidence table; the further-reading misattribution; no mention of the displacement-current error.
• example.py. Runs cleanly with .venv/bin/python. Its output (3699.2 / 3667.3 m/s; v = c for any pair mass; 1/α miss of 2.223e-6, or 13,892σ) matches the primer and arbiter.

6. Remediation actions

Deep reads (most important first)

1. TheGeometryofParticlesandtheExplanationofTheirCreationandDecayv2-ac6e61, pp. 2-22, Opus. Read every equation as a page image. - How is E_ν = 2.39 eV obtained (Eqs. 2.1-2.10, the "g-factor")? Is any step an identity or a tuned constant? - What K values does Fig. 4.2 assign to every particle, from the neutrino to the Higgs? Is any K or mass predicted before comparison? - Confirm Table 5.1 and the K = 28 vs K = 29 choice. - How do the composites (proton = 4e⁻ + 1e⁺, neutron = 5e⁻ + 1e⁺) get their charge, spin and mass? - Is the decay mechanism (solar (anti)neutrino triggers) stated as testable?
2. TheGeometryofSpacetimeandtheUnificationofForcesv2-3-f01483, pp. 10-32, Opus. Page images of sections 3.2-5.9. - For each "derived" constant (Bohr radius, Rydberg, h, Compton wavelength, N_A, Bohr magneton, fundamental wavelength 2.7e-17 m), is it an identity of the inputs or a genuine prediction? - Does the geometric α (off by 2.2e-6) feed into the other outputs? - What exactly is claimed about the "hydrogen to calcium" validation and its reference [15]?
3. ParticlesoftheUniverseMeetsElectricUniversev2-37c6b2, pp. 5-13, Sonnet. - Are the strong force (42,127 N), the inverse-cube orbital force at the Bohr radius, and the Rydberg/h derivation with Δ identities? - What is claimed for multi-electron atoms?
4. PlanckConstantAndTheNatureOfLight-V2-db51e8, pp. 6-10, Sonnet. - Read Appendix A (proton "force" constant) and Appendix B (displacement current, camelback effect). - Compare with V3 pp. 8-10. What exactly is claimed about ∇×H, and is the claim kept in V3?
5. CNPS-2018-ProceedingsJeffYeeLoriGardi-026535, pp. 4-9, Sonnet. - Are Distinti's New Induction and New Magnetism "equivalences" and the gravity "three ways = 1.97e-56 J" identities? - Is any testable difference from standard equations stated?
6. PaperColour-6aab39, pp. 5-12, Sonnet. - What precisely is predicted about pear-shaped photon rings and the singularity position? - What are the M87 "22 light-year offset" and 3C-186 alignment claims? - What do the S-star and bubble-chamber comparisons claim, and is any of it quantitative?

Changes to later files

• facts.json
• Geometry of Particles: correct how_read, summary and key_claims to include Eq. 4.10, Table 5.1, the fitted E_ν, the composite proton/neutron and the solar-neutrino decay trigger.
• Electric Universe: add the strong-force, orbital-force and Rydberg/h claims.
• Planck V2/V3: add the displacement-current claim, and correct the V2 how_read once the body is compared.
• CNPS 2018: add the gravity-identity claim.
• Web-check the KATRIN bound and the PDG muon and tau masses, and retag them [web].
• claims.md
• Claim 6: add "lepton masses follow E = E_ν K⁵ Σ…, fitted to the electron; neutrino ≈ 2.39 eV; proton = 4e⁻ + 1e⁺".
• Claim 4 or 7: add "Maxwell's displacement current is spurious (∇×H = J)".
• Source gap: add Yee's atomic-orbitals paper (ref [15] of Spacetime) to the companion works not read.
• Make the page-image page numbers (PDF vs printed) consistent.
• arbiter.md: add a "QE revision" section that:
• (a) changes claim 6 (particle model) from "Never properly tested" to Tested, not supported: neutrino-mass outputs contradicted by KATRIN and cosmology; charged-lepton fits post-hoc with about 19 % K-spacing; muon off by 10 % with K = 28 chosen over a better K = 29.
• (b) adds Internally inconsistent for the composite-proton charge count, unless the deep read finds a stated rule.
• (c) adds the displacement-current removal as Refuted by calculation (charge continuity and the vacuum wave equation).
• (d) closes the MUA reflector question: the paper's own baseline uses P/c for a perfect reflector, which is an error.
• (e) updates claim 3 with whatever deep reads 2-3 find.
• primer.md
• Fix the claim 6 row of the evidence table and section 7 ("What is genuinely open" no longer applies to particle masses).
• Add a short worked check of Eq. 4.10, with the K-spacing argument, to section 5 or 7. This is an excellent learner exercise on post-hoc integer fitting. Add a matching check-yourself question.
• Fix the further-reading attribution (Electric Universe = Yee and Gardi; electrigen = MyART, Dinu and Gardi).
• Mention the displacement-current error in section 4.
• Optionally note the 1930s "electrons in the nucleus" history.
• example.py: optionally add a section G that reproduces Table 5.1 and E(K+1)/E(K).
• contrarian.md / physicist.md / mathematician.md: no rewrite needed under MINOR. The arbiter's QE revision supersedes their claim 6 sections. Fixing the contrarian's 4e-16 → 2e-15 slip is optional.

---

Recheck

Recheck run 2026-10-03 (Sonnet). Read: arbiter.md (QE revision section), claims.md, facts.json (valid JSON), primer.md, example.py (run in the project venv), deep/ (six reads present). Source folders not touched.

Defects

Remediation actions

Remaining issues

1. facts.json still tags the KATRIN bound and the PDG muon/tau masses as [memory] with "web check pending", although the arbiter and primer have the [web] checks. This is a consistency gap only, with no effect on any verdict.

Final: OPEN (1 issues)

---

Recheck 2

Recheck 2 run 2026-10-03. Re-read facts.json (valid JSON), claims.md, arbiter.md, primer.md and deep/ (six reads present).

1. facts.json still tags the KATRIN bound and the PDG muon/tau masses as [memory] with "web check pending": Fixed. facts.json lines 303-304 now carry [web] tags: muon 105.6583755(23) MeV and tau 1776.86(12) MeV (PDG values via Wikipedia), and KATRIN m_nu < 0.45 eV at 90% CL (arxiv 2406.13516). This matches the arbiter and primer. Only the old 0.19 MeV and 18.2 MeV neutrino limits remain tagged [memory] (line 305), which is correct for a historical value that does not bear on the verdict.

Final: PASS