Bingo. Not enough units yet to make a definitive assessment but you are bang on; the preliminary numbers aren't looking good.Isn't it 'convenient' that the variance has not swung in the other direction even once just yet?
Inter-unit variance or what you are calling it should swing both ways, then I am sure we would all just admit that's the nature of things.
We have three 9500 samples and three 9900 samples.
My 9500 has 800 hours on it. Woofers has 450 hours, yours has how many?
In all three cases, the 9900 was the dimmer projector by a minimum of 10%.
Woofers would actually be the worst, since he measured 22% BUT did not take into account a known reading when the 9500 was new, so accounting for that, sounds like his is the dimmest of the lot.
Mine is ~17% accounting for known 9500 OOTB measurements to compare them to.