Your system description covers two different configurations, and appears not to have been updated in many years. Are you presently using biwired Goertz MI-2 speaker cables, as shown in one of the descriptions? And if so, are you using them without a Zobel network?
If so, between the ultra-high 950 pf/foot capacitance of those cables, which as seen by the amplifier would be doubled in the biwired configuration, and the very low impedances and highly capacitive phase angles which I presume your M-L Odyssey speakers have at high frequencies, you would be subjecting the amplifier to brutally capacitive load conditions at high frequencies. While normally I would be among the last to suggest that the symptoms you are describing be addressed by focusing on cables, if the answers to my questions above are "yes" it seems very conceivable to me that the effects of all that capacitance on the amplifier could be a significant contributor to the problem, exaggerating the consequences of excess high frequency energy when it is present in the recordings.
As far as digital EQ is concerned, you may also want to consider the DSPeaker Anti-Mode 2.0 Dual Core, at around $1100, which has been getting a lot of good press. (I have no experience with it). Among many other functions, as described in
Kal Rubinson's review (scroll down to the middle of the page) it "includes a 16-band, user-configurable parametric equalizer with a center frequency range from 20Hz to 24kHz, each band assignable to the right, left, or both channels."
Based on reviews and comments I have seen its transparency is quite good, but if you go that route I would, as suggested by some of the others above, still configure the system so that it could readily be switched completely out of the signal path when desired.
Regards,
-- Al