ONLY real problems is using unbalanced 1/4" inputs being fed with mics, but will work again to be low noise being driven from unbalanced low output impedance preamplifiers with noise almost identical to graph below showing balanced preamplifier noise plot into TRS inputs.

damn that 1/4 unbalanced mic input plot is ugly!! 
ONLY real problems is using unbalanced 1/4" inputs being fed with mics, but will work again to be low noise being driven from unbalanced low output impedance preamplifiers with noise almost identical to graph below showing balanced preamplifier noise plot into TRS inputs.
just wondering, why would you get high noise with a mic feed but low noise from an unbalanced low output impedance preamp? Thanks for sharing your findings. +++
Most mics have higher >300 ohm impedance than typical ~50 ohm drive ability of solid-state preamplifier output. With preamp driving unbalanced input, the ground reference noise is coupled into both -/+ TRS input in nearly equal manner, so noise is canceled (common-mode) in same way as balanced input. With higher impedance like from mic input, the noise signal is strong going into -minus input, and less strong (due to mic output ohms) into + plus ohms, so there's far less common-mode rejection of the noise signal riding on the input ground reference.
Got word yesterday from M-audio tech my sent-in-for-repair or refund is
NOT repairable for known issues, and purchase price for MT2 is being refunded.
So
for sure unbalanced input noise issue is now officially an MT2 'product feature' as explained by M-Audio.
However, although not mentioned, the 88.2K bandwidth shortcoming with MT2 applying wrong filter
may still be firmware corrected, but I'll never know for not getting the unit back to bench test.
I hope for owners of MT2 using 88.2K mode
at least this issue is not a 'standard feature' and will eventually be fixed with a firmware update. So those using 88.2K sampling rate might still have a non-addressed problem with ~ 20,000 cycle bandwidth limitation until further notice.
So I continue to use my MT1 as the better Microtrack model version for analog input performance reasons.