We propose a protocol-level reinterpretation of gravitational sourcing in a dynamic-update universe. Building on empirical observations showing a systematic scaling of the Ornstein–Uhlenbeck correlation time τ_c inferred from timing residuals across multiple spacecraft platforms (GPS, Juno, Cassini, New Horizons) and heliocentric distances (1–20 AU), we treat τ_c as a synchronization relaxation time—an observable measure of consistency-recovery latency in what we call the "synchronization mesh": a conceptual framework in which spacetime is continuously updated through local interactions that propagate at finite speed. Core Observational Axiom: τ_c = κ √R, κ ≃ 91 s/√AU This is equivalently expressed as a diffusion-type recovery law: τ_c² = λ (R/c), λ ≡ κ²c suggesting that synchronization recovery behaves as a random-walk accumulation over the causal propagation time R/c. Theoretical Mapping: (i) The speed of light c is not treated as a primitive "cause," but as the implementation limit of synchronization propagation. (ii) Curvature is interpreted as the geometric shadow of synchronization cost. (iii) What is conventionally called "dark" (DM/DE) may be reframed as the accumulated, non-disposable overhead of continued global synchronization and space updating.

V112: √R Diffusion Law — Observational foundation (companion paper) V113: Synchronization Mesh Protocol — Formal specification (companion paper) V108: N4 Minimal Structure — Theoretical foundation (companion paper) V110: N4-Macro Bridge — Scale transformation (companion paper) Supplementary Material Added (2026-02-12): Appendix E — "Inner Solar System Verification Report (Rosetta-Integrated Edition)" This appendix documents the observational evidence supporting the τ_c = κ√R scaling claimed in the main text. Key contents: (1) Rosetta RSI multi-band verification — X/S-band agreement within 2% across all detrend orders, excluding dispersive plasma origin by a factor >13; (2) IFMS1/IFMS2 hardware reproducibility at 15× suppression, consistent with round-trip phase cancellation prediction; (5) Integrated 13-point verification table spanning 0.36–20 AU (9 spacecraft, 4 agencies, 17 years), yielding κ = 94 ± 12 s/√AU with mean agreement 90%. Full systematic error budget, detrend protocol, and predictions for archival datasets (MaRS, Galileo, Cassini, Voyager) are included.