Parameter-Penalized Real-SPARC Diagnostics of the VES Galactic Response From RAR-Level Field-Response Surrogates to Halo Baseline Competition

This work presents a staged diagnostic benchmark of Viscous Emergent Spacetime (VES) field-response models against SPARC galaxy rotation-curve data. The analysis uses 3,391 rotation-curve points from 175 disk galaxies and compares baryon-only predictions, a fixed RAR/MOND-like baseline, three VES 6D variants, and standard two-parameter NFW and Burkert halo fits. The final Stage 6F-VES benchmark performs a per-galaxy parameter-penalized comparison using RMSE, weighted chi-square, AIC, and BIC, with an observational velocity-error floor of 1 km/s. VES 6D-1, VES 6D-2A, and VES 6D-2B are treated as zero per-galaxy parameter response models, while NFW and Burkert are penalized as two-parameter per-galaxy halo fits. Strict point-level merge validation confirmed that all three VES prediction tables matched the full benchmark set: 3,391 rows and 175 galaxies for VES 6D-1, VES 6D-2A, and VES 6D-2B. The median BIC ranking was Burkert 14.387, NFW 21.707, VES 6D-1 140.299, VES 6D-2A 147.166, VES 6D-2B 148.041, fixed RAR 154.490, and baryon-only 1431.693. Thus, all tested VES variants improved over the fixed RAR baseline by median BIC, with VES 6D-1 giving the strongest VES score. However, the two-parameter halo baselines remained substantially preferred, with Burkert and NFW dominating AIC/BIC best-model counts. The result should be interpreted conservatively. VES 6D has reached a RAR-competitive regime and slightly surpasses fixed RAR in this diagnostic setting, but it is not yet competitive with independently fitted halo models. The benchmark therefore motivates the next research stages: variable stellar mass-to-light testing, one-parameter or hierarchical VES amplitude extensions, residual-structure diagnostics, and replacement of the surrogate disk-response kernels with a full elliptic field solver.