=== (1) SiC thickness vs admissible group index ===
4H-SiC Sellmeier at 2900 cm^-1: n = 2.5657, n_g = 2.5707 (n_g - n = +0.0049)
   n_g   d(10deg)   d(15deg)   vs ref 7.604
 2.550     7.6919     7.6715         +1.16%
 2.570     7.6320     7.6118         +0.37%
 2.575     7.6172     7.5970         +0.17%
 2.600     7.5439     7.5239         -0.79%
  => over the physically admissible SiC group index (2.55-2.60), d stays within ~1.2% of 7.604.
     A model implying n_g = 2.67 (n_g - n = 0.20, ~40x the physical dispersion) is not admissible.

=== (2) multi-beam detection: second-harmonic energy ratio ===
  SiC_10deg.xlsx E2/E1 = 0.0005
  SiC_15deg.xlsx E2/E1 = 0.0044
  Si_10deg.xlsx  E2/E1 = 0.9121
  Si_15deg.xlsx  E2/E1 = 0.8935
  => SiC ratios are ~1e-3 (negligible multi-beam); Si ratios are ~0.9 (strong multi-beam).
     A threshold near 0.02-0.05 separates them by two orders of magnitude.
     SiC: two-beam model, no correction needed. Si: Airy multi-beam model required.
