Fighting with Flats on the QHY-9

I've yet to figure out the issue with my Flats.  I take them with the same methodology as I did with my QHY-8 OSC in which, I had no issues.  I use either the "white tshirt" method, or the "dusk/dawn sky" method.  My color channels are not aligned at all regardless of what I do.  I can crank the gain to 100, 50, 0, no change.  Amp on, Amp off, no change.  Offset adjustments don't affect the seperation.  Exposure time changes nothing but where the three peaks show up in the histogram but they are always seperated.  I'll add that the optimal exposure times to land around 25K ADU/Max readout is at least several seconds in dim or strong light so the time is high enough that I'm thinking the shutter is not a factor here, I think?  I've checked for light leaks in the imaging train, and just to be sure I've wrapped everything up, except the fan on the camera of course.  I've had TEC on and off, no difference there.  I can't think of what the issue is.  However, I've yet to apply Flats with this camera that didn't worsen my light frames significantly.  Usually creating a nasty color gradient across the image - an effect that isn't present without the Flats.  Unfortunately, vignetting and dust motes are left behind of course.  I just don't get the problem here.  The IR filter is perfectly clean.  It's not in an extreme hot or cold environment right now.  I don't know what to do here.  I've attached a screenshot of my bias frames and data around it, along with a sample Flat, and the same Flat debayered as well so that I can show the color value readout.  One thing I'll note - when I debayer, then auto color balance, the flat looks like a proper flat.  Ideas?  I'm out of them.