# n237 — Second anomaly sweep: the CODATA-G puzzle adjudicated from within, the disorder scale bracketed by two anomalies, and the refuse list

**Date:** 2026-07-12 · **Directive:** look again at the anomalies. The productive find: a famous metrology anomaly turns out to have exactly the shape of one of the glass's native structures — and testing whether the theory can claim it *bracketed a parameter instead*.

## 1. The CODATA-G discrepancy: the shape matches, the scale is forbidden — the systematics bet becomes a derivation

The known anomaly: G experiments are individually precise (10–30 ppm) yet mutually discrepant (~500 ppm) — a real, unresolved metrology scandal. The glass's frozen force-scintillation (n225) predicts *exactly this pattern*: each apparatus samples its local, disorder-frozen G — locally repeatable, globally scattered. The temptation is obvious. The arithmetic kills it, beautifully:

- To produce a 5×10⁻⁴ spread at apparatus scale, the vacuum's disorder correlation length must be **ℓ_d ~ 0.6 mm**.
- But GW coherence forbids it: at ℓ_d = 0.6 mm, the random-walk phase scatter over GW170817's 40 Mpc is **~10³ rad — waveforms would be destroyed**; LIGO's clean chirps exclude mm-scale vacuum disorder outright.
- At the atomic-class ℓ_d the theory actually leans toward (a₀), the GW phase scatter is safe (0.3 rad) and lab-scale G-scintillation is **10⁻¹⁴** — eleven orders below the CODATA spread.

**Verdict: the theory cannot absorb the G anomaly, and therefore prediction 13's "it's systematics" bet is now *derived from within*, not just declared.** Bonus structure: two known observations now bracket the vacuum's disorder correlation length from both sides — **GW coherence caps ℓ_d ≪ mm from above; the scintillation mechanism plus any future confirmed G-inconsistency would probe it from below.** A parameter the theory could not price got bracketed by anomalies instead.

## 2. Flagged direction [C], not claimed: the low-ℓ CMB anomalies

The hemispherical asymmetry and quadrupole–octopole alignment (2–3σ, persistent) have a natural home in a KZ-crystallized vacuum: the quench correlation length at freezing is a preferred scale, and defect-seeded super-horizon modes are generic KZ products. No FRW sector exists to compute it — registered as the cosmology sector's most natural future target, with the honest note that 2–3σ CMB anomalies have died before.

## 3. Swept and refused (F10 hygiene)

Muon g-2, proton radius, lithium, EDGES-21cm, kaon anomalies — matter/QED sector absent: refused. Flyby anomalies — dead data (Juno nulls) + scintillation self-averages at planetary scale: refused. AU-drift/lunar eccentricity — inter-ledger flow is capped by the same clock/LLR bounds as n231: refused. S8/growth tension — no mechanism: refused. JWST early massive galaxies — no FRW: refused. Quasar-spin alignments — the glass has no coherent axis; these must dissolve (exposure, joins entry 15's family).

**Class: §1 [A — arithmetic on measured structures; the bracket is the result]; §2 [C, flagged]; §3 [refusals]. The sweep's lesson: the theory is now constrained enough that famous anomalies get *verdicts* — claimed, bracketed, or refused — rather than speculative embraces.**