Update figures and docs: OMA naming, uniform legends, TikZ Fig. 1 source

- Rename the orthogonal comparison scheme OFDMA -> OMA in docs and figure legends (data keys keep ofdma)
- Legend labels normalized to the paper dictionary (OMA, SFDMA, UWCA (proposed))
- New TikZ-generated system overview figure with text-consistent notation (soft decoder masks m-tilde, queries q_u, sharpness eta)
- Regenerated fig_fair, fig_resorth, fig_realdata_c, fig_async, fig_dynusers; fig_ser_all legend patched

README.md

@@ -70,7 +70,7 @@ Each study script is self-contained and writes its JSON into
 inequality in the manuscript with standalone numpy code (no experiment code
 reused): the relevance identity, the mutual-information correlation, the
 subspace ceiling, the LMMSE receiver against an empirical Wiener solution
-(0.1% MSE agreement; the blind form matches the OFDMA cosine to machine
+(0.1% MSE agreement; the blind form matches the OMA cosine to machine
 precision), and the surrogate bound used in the appendix (uniform constant
 0.97). Results: `experiments/verification/math_verify.json`.