Just to clarify:
-
Normally duv and CRI are independent (ok, not technically true because 100CRI is 0duv by definition), but in this case they are related because of a common underlying cause: an excess of blue emission, which simultaneously drops duv and CRI.
-
The harshness I see has nothing to do with brightness, I refer to the spectrum of the light. Reds and pinks in particular become more blue shifted and desaturated, and white colors look off-white.
-
The drop in CRI as current increases is a known phenomenon, a recent example is the 719A test. Looking at the 5000K variant, the CRI drops from 92 (68 R9) at 0.3A to 88 (38 R9) at 3A, accompanied by a growing blue spike in the spectrum.