Sure. Clearly you have strong beliefs about this and since this is specific thread I'll close with these:
It's not just me, most experts in the field appreciate the need for spectral balance. Even Dr. Olive who is promoting a room correction approach with a tilted response, still understands the benefit of using speakers that are spectrally balanced when listening at reference.
|1) Few regularly listens at reference level so setting that as a contraint isn't helpful for most of us. But let's say it's 10% of the time. Then 90% of the time flat response is wrong.
2) Reference levels (and EQ) for non-standard mixes are generally unknown in any case.
For situations where the reference level is unknown, Audyssey permits us to experiment with the Reference Level Offset values as an approximation to get us in the ballpark. Other more precise approaches can also be used even if an AVR does not have the Offset feature. One can experiment with the Input Level settings (which have finer adjustments), up or down, to readjust the amount that Dynamic EQ tilts the "flat" response initially derived by MultEQ. Dr. Olive's experiments demonstrate that someone with a trained ear can hear spectral balance even if the recording's reference level is unknown. (I don't pretend to be one of those folks. ) So these folks could adjust the Input levels to permit Dynamic EQ to accurately restore spectral balance.