/ how have people been modeling mismatch? monte carlo methods are only useful if you know the underlying parameter distribution.

Monte-Carlo can be used but it is a long way round. The issue is that monte-Carlo doesn't work either. @Tim Edwards is there any way we can get some mismatch info from skywater? Pelgrom coefficients, measurement data etc? As you know, designing the performance of things like ADCs is kind of a shot in the dark without some data to design upon