V.R392 — Connection to Cocycle-Defect Minimization

At galactic scales (r/ℓ_τ ~ 10⁻⁵), the NF iteration operates where ε = r_g/r ~ 10⁻¹³. Convergence to linearized solution plus exponentially small corrections — cannot generate factor c²ℓ_τ/(2GM) ~ 10¹² needed. The gap is not an NF failure but a feature: the full solution encodes a₀ = c²/(2ℓ_τ) through geodesic equation, not scalar PDE.