# n109 — the first external critique: triage and response (2026-07-10)
# Received in full; the program treats it as it treats everything:
# accepted findings applied same day, disagreements argued, all here.

## The critique's verdict, quoted fairly
"Excellent site for making one want to read; bad site for correctly
calibrating the degree of proof. The main flaw is editorial: the site
knows exactly what it has not demonstrated, but keeps writing its
headlines as if it had."

## ACCEPTED AND APPLIED (same day, commit-stamped)
1. THE 14/22 INCONSISTENCY — real, and shameful for a ledger site.
   Fixed. Versioning adopted: The Audited Vacuum v0.3, date + commit
   in the footers; pages now carry their version.
2. "MEASURED" — the critique is right that the word implies
   laboratory. All treatise instances relabeled: "toy (finite
   lattice)" / "numerically observed, 12-200 sites, by the authors."
   "Measured" is henceforth reserved for external data.
3. "ONE AXIOM" — overclaim. Headline now "One principle. One model
   class." with the assumption list linked in the first paragraph.
4. EDITORIAL HIERARCHY — the front page now carries the epistemic
   legend in its opening paragraph (theorem/toy/reading/data), and
   every entry's stamps show class, not celebration. The calibration
   no longer lives only on a later page.
5. "ELECTROSTATICS"/"PHOTON" — requalified: "Coulomb-like sector,
   read as electrostatics"; the gapless mode "read as the photon,"
   with the unbuilt list (vector field, polarizations, Maxwell,
   Lorentz, charge quantization) inline.
6. "NO ADDED NOISE" — now scoped: "in the computed observable of this
   finite Gaussian model." The amplifier-limit point is well taken.
7. "REDSHIFT CONSTITUTED BY QM" — softened to the defensible claim:
   in this model class, energy-derived phases read the potential
   where our classical observables read gradients; GR's classical
   proper-time account acknowledged; "gravity was never classical"
   demoted to a hypothesis about nature.
8. MERCURY-LAMBDA — headline changed to "one anharmonicity could feed
   both"; the missing steps (T_munu form, scale, homogeneity,
   covariance) named inline.
9. THE RIVALS TABLE — corrected: graviton-induced decoherence master
   equations exist; colored-noise CSL has frequency structure; the
   table now says "compare to defined models, not silhouettes."
10. "A FLAT NULL KILLS EVERYTHING" — restated at true strength: a
    shape prediction without an absolute rate; a null excludes what
    its sensitivity reaches.
11. PREDICTION HIERARCHY — primary discriminators (the knee family)
    now distinguished from corollaries and stances on the predictions
    page.
12. METAPHOR DENSITY — reduced where claims were carried by rhetoric
    ("owns the whole house" -> "carries the falsifiability", etc.).
    The audit vocabulary retained where it names mechanisms.

## ACCEPTED AS PROGRAM DIRECTION (already the named organs)
The critique's three deliverables are the program's own list, and we
co-sign the prioritization: (i) a microscopic model of the active
reservoir (the pump) with energy budget, steady state, fluctuations,
stability, and the universality of its coupling — "a demon needs a
battery" is an intuition, not a derivation, exactly as charged;
(ii) a relativistic continuum limit with identified spin content;
(iii) an absolute knee-rate prediction with required sensitivity and
exclusion regions. Until then, the corollary predictions are
consistency tests, so labeled.

## ARGUED (two pushbacks, narrow)
A. "The site's discipline is rare and its Status page excellent, but
   too late" — we note the same review found the graveyard, the
   settled-against-us forecast, and the self-demotions BEFORE we
   fixed the hierarchy: the record's honesty was load-bearing even
   when the front page oversold. The fix was owed; the substance was
   already there.
B. "22 predictions dilute" — partially resisted: the corollaries
   (antimatter nulls, LIGO scorecard, proton silence) are how a
   framework shows it does NOT contradict existing precision data;
   removing them would hide passed checks. The fix is hierarchy (now
   applied), not deletion.

## What this critique is, in the record's terms
The first genuinely external eye — and it performed exactly the
function the program reserved for it: it found the inconsistency our
own sweeps missed (14/22), named the systematic bias our own
contradictory analysis (n85-F8) predicted ("same-agent registration
is not a firewall"), and forced the calibration from the back page to
the front. The site it criticized sold a nearly-complete theory; the
site it leaves behind sells a model class with one great experiment.
That is the correct product.

## SECOND READING (same reviewer, same day) — triage
CONFIRMED FIXED (reviewer read a cached page): the 14/22 count.
NEW BUGS FOUND AND FIXED:
- "six equations, four mechanisms" title vs five Mechanism sections
  (stale since Mechanism 5 was added) -> retitled.
- Prediction ordering 1-11, 15-22, then 12-14 (entries were inserted
  before the wrong block) -> reordered.
- The window/anchor tension (band 1e19-1e22 vs anchor 5.7e18):
  explained inline — the band was the order-of-magnitude bracket; the
  G-cancellation sharpened its electron edge to 5.7e18; the anchor
  sits at the refined lower edge. And the anchor is now labeled what
  it is: A HYPOTHESIS (the bound-saturating electron as counting
  unit), adjudicated by the experiment, not derived.
NEW OVERCLAIMS FIXED:
- Index's "exact Coulomb in 1, 2 and 3 dimensions" -> 1D exact fit;
  2D/3D are dispersion checks, said so.
- Results' theorem label "photon masslessness" -> "gapless Gauss
  mode"; the name "photon" explicitly a reading with its unbuilt list.
- The flat-null claim, final form: a null excludes reached amplitudes
  only; TOTAL falsification awaits the absolute rate — THE
  FALSIFIABILITY GAP, now named on the experiment page as the third
  priority deliverable. The reviewer's demand for a nonzero amplitude
  floor is accepted as the gap's definition.
- Corollary predictions (15-22 class) explicitly graded: structural
  extrapolations into unbuilt sectors, weakest class on the site.
The reviewer's closing formulation is adopted as the site's own:
"a class of mathematical-physics mechanisms, verified on small
models, whose transfer to nature remains a hypothesis."