ARTICLE2 SR OBS-ACT EXTENSION — 2026-07-20

STEADY CALIBRATION
  Constant rear counter-rotation sweep (alpha magnitude 0..6): best sampled alpha=5,
  legacy DTW=0.987212. Joint Karman formula in steady inflow: DTW=0.966494;
  rear tail action=(-3.4306,+3.4318), analytically equal to constant magnitude 3.4312.

400-STEP TRAINING-CONDITION VALIDATION
  Karman: Re50 0.952178; Re100 0.944339; Re200 0.863850; Re400 0.833365.
  Illusion: 0.75L 0.854379; 1L 0.915609; 1.5L 0.833587.
  All completed with finite telemetry. These values use legacy_dtw_v1_abs_n_unclipped.

CURRENT-FORMULA GENERALIZATION (200 STEPS)
  Karman: Re25 0.988078 (extrapolation), Re70 0.954339, Re150 0.907496,
  Re300 0.835758 (interpolation).
  Illusion: 0.5L 0.784055, 0.6L 0.821103, 2L 0.756880 (extrapolation);
  0.8L 0.884573, 1.2L 0.925530 (interpolation).

PER-CASE DIAGNOSTICS
  Karman rear constants remain O(3-4) but gains are non-monotonic; no clean low/high-Re
  two-group law is supported. Illusion 1.5L has a distinct coefficient emphasis, while
  single-case rear refits at 0.75L/1L are poorly identified and are not controllers.

HANDOVER
  Highest-priority OID/CCD targets: rear steady counter-rotation; rear symmetric lift
  feedback; front lift in Illusion; rear drag asymmetry. See sr_key_elements.md/csv.

LIMITS
  No direct momentum-balance equality is claimed. Per-case refits are offline diagnostics.
  Generalization claims are limited to the sampled conditions. OID/CCD causal/spatial
  structures remain to be established.
