GCFT CDN-166: The Velocity Window — Time-Domain Drag Decides the vᵣ-Linearity Question

CDN-166 settles theory debt D2 from the CDN-164 ledger. A time-domain integration of the CDN-163 linearized spintessence system (RK4, spectral Laplacian, absorbing sponge boundaries) yields a true drag force on a moving gravitational source — sign opposing motion, converged under domain size, run length, and timestep. The drag is linear in velocity only for v ≲ 0. 1 (F/v constant to 7%), peaks at v* ≈ 0. 12, and collapses by a factor ~500 by v = 0. 6 (local exponent ~−7): the fast source outruns the gapless quadratic branch that carries the wake. Verdict: the vᵣ-linearity assumed since CDN-139 is a low-velocity limit, not a law; the coupling must be read as acoh = −κc·f (ρ) ·W (vᵣ/v*) ·vᵣ/r² with a window function W. All κc-based results acquire a velocity-window caveat, and v* in physical units enters the ledger as new debt D5. Labeled conjecture: if v* falls between the MW–M31 approach (~110 km/s) and typical satellite velocities (~300 km/s), fast systems decouple kinematically — relieving the stream tension while preserving the CDN-139 flyby signal — and predicting a velocity-sorted anomaly pattern in the Gaia DR4 satellite sample, distinct from pure density screening. Channel split versus velocity matches CDN-163/165. Solver, results JSON, and figure included. By Nicky Hacquier (Aporion Dynamics) — Project Aporion; follows CDN-165 (DOI 10. 5281/zenodo. 20634700).