Gate RM3 MoorDyn sea and portable MOST coupling balance

[cmccomb]

What changed

• Rebuild the published 80 s RM3 MoorDyn incident sea in ordinary CI from archived MATLAB phases and the bundled RM3 hydrodynamic data. The independent elevation gate is 2e-14 m.
• Exercise the real MOST BEM, controller, rotor, tower, mooring, and platform together on a short synthetic case. Check initial heave acceleration against an independently calculated force/mass balance and compare causal with converged coupling.
• Clarify that source-step-sensitive cases remain qualified without a step-stable physical reference; badge case counts are unchanged.

Verification

• Focused tests: 5 passed locally.
• Full Python suite: 254 passed, 333 MATLAB-dependent tests skipped.
• Combined line-and-branch coverage: 75% to 76%. The coverage badge reflects the measured result.

The prior master MATLAB sweep's RM3 PTO Extension job failed during MATLAB installation, before any WEC-Sim simulation; two other jobs were cancelled. The last completed full sweep passed. This PR does not change source workflow or numerical thresholds.