Beyond the Coherent Limit: Finite-Response Corrections and Metric-Closure Failure in ECSM

This paper develops the finite-response layer of the Emergent Condensate Superfluid Medium (ECSM) framework beyond the coherent general-relativity limit. In ECSM, general relativity is recovered when the underlying medium response is coherent, long-wavelength, and locked across clocks, inertial motion, optical propagation, and tensor disturbances. This paper asks what happens when that metric closure begins to fail. A finite-capacity response closure is introduced, allowing different physical sectors — dynamics, optics, lensing, CMB-scale response, and laboratory decoherence — to sample different coherence states while remaining tied to one underlying finite-response architecture. The paper then proposes an effective action with response-dependent curvature coupling, a dynamical coherence field, and a finite-response kernel. Variation of this action gives a structured correction tensor that vanishes in the coherent limit. The main proposed observable consequence is a correlated lensing-dynamics split: in coherent regimes, lensing and dynamical gravitational response agree, while in finite-response regimes photons and slow matter may sample different coherence states. This gives a falsifiable target for strong lensing, weak lensing, galaxy kinematics, cluster dynamics, and environmental dependence. The paper also introduces a scale-closure ledger linking the galactic acceleration scale, optical saturation, lensing response, and drive-dependent decoherence to a shared finite-response parameter structure. It does not claim to be a completed microscopic theory, but presents a disciplined bridge from coherent GR recovery to quantitative finite-response phenomenology.