I think I have found a solution for the audio that will work, although I still don't know if this is the best solution.
I leave the external stereo audio as one track on the pre-nested sequences. In the multicam sequence I can drop the Fill Left or Fill Right effect onto the audio clip depending on who is talking. If I need to pan audio, I cut the clip and put the opposite Fill effect on the next clip and put an audio transition between them.
It still doesn't solve the problem of having to go back to the original sequences in order to use the Motion controls to reframe. This is indeed fairly tedious as it involves finding the point in the multicam sequence that I want reframed, pressing M, copying the timecode in the source window, going back to the pre-nested sequence, pasting the timecode in the timeline window, making a cut, adjusting scale and position. And possible finding the out point and reverting to the default.
Does this sound like madness?
When shooting weddings, we will have from three to six video tracks and a half-dozen audio tracks. I'll skip the explanation of getting six v tracks boiled down to four, since it is irrelevant to your question..
I take the audio file with the most information (i.e.: enough sound to figure out what is going on) and place it on track #1 in my source sequence. When I'm finished cutting the multi-track source video, and before cutting out any of the target sequence video, I turn off audio track #1 in the target sequence (or just delete the track) and copy all the audio tracks from the source sequence into the target sequence. Any stereo tracks that had different mics, I'll copy twice and use left or right fill as you've done, to isolate the mics. Often I'll rubberband the audio tracks so I don't pickup from the further mics, etc.
With only two cameras, you may be just as well off laying one video track over the other and cutting away the upper track as needed ala PP 6.5, but when using three or four cameras, the multi-cam feature sure comes in handy.
Don't forget to sweeten your audio a bit with the compression feature of the dynamic processing options.