Iz
The primary problem caused by high IMD is splatter outside the bandpass of the sideband filter. This shows up on the opposite sideband and on the high (audio) frequency side of the desired sideband. We have all heard this when an overmodulated signal is parked 3 KHz away and we get "static crashes" in the 3 KHz that we are listening to. On a waterfall is shows up as a broadening of the base of the signal.
I don't think you will find the IMD products affecting the intelligibility of the audio even at the very high -12 dBc levels I measured. What did happen is the transmitted noise floor on the opposite sideband and outside the filter rose 20-30 dB over the the single tone level. If there is interest I can replicate that test and show what it looks like on the spectrum analyzer.
WA8TOD
