The Universal Law of Coherence: A Consolidated Framework and Empirical Validation

This version (V5) extends preregistered testing of the Law of Coherence (LoC) to additional domains (GR ringdown and NMR FID). Key outcomes: Endurance supported: In all valid datasets, Δ > 0 versus strict IAAFT surrogates, confirming LoC’s necessary condition. Scaling suggestive but inconclusive: The pooled Δ vs ln(τ) regression shows a negative slope (≈ –0.3) with moderate explanatory power (R² ≈ 0.49). However, the 95% confidence interval crosses zero and permutation testing does not reject the null (p ≈ 0.26). With only 4 valid rows, the analysis is underpowered. Contents: Public package — Δ and endurance results only (minimal, for readers). Archival package — full reproducibility metadata (logs, checksums, environment). Unified addendum — pooled Δ vs ln(τ) regression results, 95% CI, permutation p, CV-R², and figure. Scope: V5 confirms endurance across domains under preregistration, but pooled scaling remains underpowered. Earlier V3 suggested positive scaling trends under looser preregistration; V5 tightens definitions, demonstrating robustness but leaving scaling unresolved.