The Boundary Prediction: Observable Consequences of Virial-Radius Density Screening in GCFT [CDN-143]

CDN-141R established density screening as the only surviving resolution of the GCFT coherence coupling scale tension, placing the coupling activation radius at 260–387 kpc — coincident with the Milky Way virial boundary. This paper converts that result into four quantified predictions supported by Monte Carlo simulation and direct NFW density calculation. P1 (Satellite Boundary): Orbital integrations of N = 2,000 satellite trajectories per initial radius show a −38% pericentre shift at r₀ = 700 kpc relative to ΛCDM, declining sharply to zero inside r₂₀₀. The transition is centred on the virial boundary and has no ΛCDM analogue. P2 (Pulsar Null): The NFW halo density exceeds ρ_crit everywhere within r₂₀₀. All Milky Way pulsars — disk or halo, at any Galactic height — are fully screened. The prediction is a clean null: any GCFT timing residual detected in any MW pulsar immediately falsifies density screening as the resolution mechanism. P3 (GW Environment): Local density spans six orders of magnitude across LIGO event host environments. Approximately 28% of current O4 events (~56 of 200) are in environments where f > 0.5 and coherence coupling would be detectable. A density-stratified rate comparison requires no new detections. P4 (LG Timing Bias): GCFT coupling during the wide-separation phase of the MW–M31 orbit biases the Local Group timing-argument mass by ~49% upward. The direction is robust: coupling adds energy to the wide-separation orbit, inflating the inferred turnaround radius and hence the inferred mass. Five explicit kill conditions are stated. All four tests are achievable with existing or near-term data.