[soundsculpter]

Thanks for your input, Mike. You hit the nail on the head as far as lack of breathing room, at least on the shows I'm mixing. Everything is wall to wall with a lot of overlap. As I mentioned before, most of my volume automation is ramping the head and tail of each track, and then just minor adjustments for the dynamic changes within the song.
I've got to be honest, and this may get scoffed at, but I don't premix the fx and music and then ride the stem for dips under dialogue. I level dialogue, ride the music, and then make a pass for nats and fx. Almost always, my nats and fx tracks are just clip-gained and aren't "dipped" within the region. Very seldom in the programming I'm mixing is it necessary to do so.
I then create the undipped M&E by duplicating the music tracks, more or less flat line my automation at the sync dialogue mix level, and route it to the M&E along with the nats and fx stem. If I do by chance have ramping or dipping within clips on the nats or fx, I do the same as I do with the undipped music.
It's kind of doing it in reverse, but I don't see how one can premix music and fx without dialogue when dialogue is wall to wall in the show. For post work, I also steer clear of riding stems for the mix because sure enough there will be revisions after it was supposedly picture locked, or down the road I'll need to conform the mix for changes for whatever reason, and I like the option of creating region groups that carry my automation. Gets tricky when things need to move and most of your mix is on aux tracks.