I don't think I've ever heard ANYONE make this argument against mixed cabs. And in fact I would disagree that it's an issue and go even further and say it's actually the reverse, in that with 2 different cabs, you are going to have a large area of overlap, combined with generally at least some space on either end which only one cab occupies. What this typically does is give you a wider, and (usually) flatter system response. Which naturally makes it EASIER to EQ.
While the point you're making would seem to make sense on the face of it, in actual practice it rarely works that way. Unless you're using cabs that have been carefully designed to work together
as a system - i.e. with drivers specifically selected for the purpose, cabs built to optimize the performance of those specific drivers, and a crossover(s) developed to function optimally with all those other particular components - the sources of frequency overlap are more likely to
interfere with each other, rather than to complement each other, thereby resulting in dead spots (phase cancellations), and hot spots (phase duplications). Using unmatched cabs, the extent to which you
might experience such problems is difficult to say. It's really kind of a crap shoot.
Usually in these fundamentalist "mixing is of the devil" kind of threads, the matched cabs mafia brings up the spectre of the twin bogeymen of phase response and dispersion. These can be important components of your overall resoponse, or they might not really matter at all depending on the gear you are using and the situations you use it in.
There's nothing "fundamentalist" about the point I'm making. I'm advocating a set of best practices that can be thoroughly backed up by science. A genuinely "fundamentalist" approach would be to take a rigidly dogmatic stance against the scientific findings - based only upon stubborn ideology.
Yet, we find ourselves in further agreement than you might have supposed. For no one is making the case that unmatched cabs
never sound good - only that the results
can be unpredictable, and therefore difficult to replicate. While a sound physicist or an audio research engineer
might be able to isolate and control for the myriad factors that can predict whether or not two or more unmatched cabs will produce satisfactory results, the ordinary
musician is typically much better off by simply playing the odds, and going with a set of matched cabs.
MM