Electromagnetic Constitutive Closure of ECSM Optical Gravity: A Split-Response Bridge Between Non-Geometric Lensing and Causal Electromagnetic Propagation

This NG11R bridge paper supplements ECSM Gravity Without Curvature by supplying the corrected split electromagnetic constitutive closure for its optical/lensing sector. The non-geometric ECSM gravity spine previously represented light bending and cosmological optical transfer through an effective scalar optical index. The corrected ECSM electromagnetic sector shows that this cannot be interpreted as a common scalar multiplier of the electromagnetic action. A homogeneous common scalar changes field normalization and wave impedance, but it does not by itself alter the Maxwell propagation characteristic. Genuine propagation change requires split electric and magnetic constitutive response. The bridge rewrites ECSM optical gravity through split constitutive functions Z_E and Z_B, with n^2 = Z_E/Z_B. In the weak coherent limit this recovers the previous ECSM optical-gradient lensing formula, but no longer as a free scalar optical assumption. The NG11R benchmark notebook passed 30/30 internal tests. The common scalar channel produced max Delta n = 0 over the scan while changing impedance as Z_wave/Z_0 = 1/Z_c. The split response reproduced the benchmark far-field optical-gradient deflection with median relative difference 5.57e-3. Omitting the domain-boundary channel produced a peak residual 0.31547 near b = 6.318, showing that boundary gradients remain physical optical-response structure after the split-channel correction. A 50,000 point parameter audit found 30,981 samples satisfying the bridge conditions. The corrected ECSM chain is: medium burden and boundary gradients -> Z_E, Z_B -> n^2 = Z_E/Z_B -> lensing and propagation. This is a constitutive bridge and benchmark audit, not yet a raw SPARC/KiDS/CMB empirical refit.