All 15 merged spectra (5–100 µW) were despiked (130 cosmic-ray spikes removed), then Lorentzian-fitted in the window 1.23–1.45 eV. Every column converged to **HIGH** quality (R² = 0.887–0.984, median 0.981; SNR 7.6–24). An allometric fit $I = a\cdot P^{b}$ was applied to the intensity trend, and a trend fit to the peak-position trend.

![Power Fit]([file link removed])

**Intensity vs power (allometric, $I = a\cdot P^b$):**
- $a = 16.28 \pm 1.81$
- $b = 1.0048 \pm 0.0261$, R² = 0.995 → essentially linear (excitonic, no saturation or Auger quenching over this range)

**Peak position (xc) vs power:**
- Monotonic **blue-shift**: $x_c$ rises from 1.3206 eV (5 µW) to 1.3356 eV (100 µW), a total of ~+15 meV (~5 meV/decade)
- FWHM narrows from ~62 meV (5 µW) to ~54 meV (100 µW)

Per-power Lorentzian values (excerpt):

| Power (µW) | Intensity (counts) | xc (eV) | FWHM (meV) | R² |
|---|---|---|---|---|
| 5 | 98.8 | 1.3206 | 62.4 | 0.887 |
| 25 | 425.2 | 1.3290 | 61.7 | 0.973 |
| 50 | 818.3 | 1.3328 | 59.0 | 0.981 |
| 75* | — | — | — | — |
| 100 | 1714.6 | 1.3356 | 54.4 | 0.984 |

Full per-power table: [power_fit_results.csv]([file link removed])

RESULT: b = 1.0048 ± 0.0261