I'd be curious to learn more about this limitation, haven't heard it mentioned before. My Audyssey MultiEQ app with Denon X3500 shows the "corrected response" matching the target perfectly.
I watched an XT32 video produced by Sound United & Audyssey in which the limitation was mentioned. Essentially, they are working within the available DSP power in the receiver and prioritize bass over treble, which makes tons of sense when you factor in room reflections.
What is shown in MultiEQ is a projected response, not a measured result. Mine shows the same thing, but since you do not re-measure, it cannot show an actual result.
Here are left and right measurements showing the response in my media room of full XT32 correction vs. limiting it to 1100 Hz, which is a multiple of Schroeder in my room:
As you can see, XT32 produces a curve equally as flawed as the raw response above 1100Hz. A little bit less even, actually. If you are well-versed in PEQ, you can see a low Q boost, medium Q cut...
This is Audyssey's reference curve with midrange compensation turned off.
My point is, one cannot rely on Audyssey to hammer a very flawed speaker into a perflecly smooth FR response. It likely will not beat a Klipsch into perfect submission.
Having said that, look at the improvement under 1100Hz compared to what I measured in Pure Direct mode (bass management on, Audyssey off):
While it could be better in the SBIR range near 150Hz, it is difficult to otherwise argue with that result. It is very good considering the hardware limitations of the AVR. Nearly everything below 1100Hz is improved by Audyssey [Note the phase was inadvertently inverted on both subs and Audyssey fixed it.]
Paraphrasing Dr. Toole, if you need to correct above the range of the influence of the room, you need better speakers.
START WITH THE BEST SPEAKERS YOU CAN AFFORD!