Can you please link these reviews? I'd like to know who mistakes accurate reproduction as tainted sound, so that I can be forewarned when I come across them.

I mean do they actually prefer zero EQ? Or they simply prefer different EQ products (which)?
Selecting xover point is pretty easy peezy. In any case, it's a whole lot easier than being successfully proficient with audio test equipment, knowing how to correlate data sampled at various locations and interpret the findings, and then be able to build filters that would yield the desired results.
And in the end, it's not Audyssey's fault anyways. You can blame the receiver manufacturers for any fault of implementation.

Audyssey gives the identical codes/algorithms to every manufacturer for any given product.
The main knock on Audyssey is lack of tweaking flexibility. Otherwise, it successfully does what it was designed to do. One of the most common complaints about the lack of tweaking has to do with not being able to cut off correction at any particular high frequency, being the Schroeder transition frequency of any room.
However, I have been informed, that even with the ARC correction available in Anthem's top end ($$$$) are users finding themselves correcting for a much larger bandwidth than expected.
It is the very, very rare room that doesn't benefit from EQ, even with a large bandwidth. When I mean rare, I'm probably* talking about ceilings well over 8 ft, with a length greater than 2x the width, with a minimum of 30 treatments. I've jibber jabbered with a few people who have about 35-40 treatments, and a couple of those people
still apply XT.
What's REALLY great about XT is how affordable it can be. I got it in a $600 refurbed receiver. It's pretty hard for me to complain about setting a xover, or being able to adjust high freq cutoff, when considering the very modest amount of money spent towards RC.