I've seen a demo of resyncing lips movement to dubbed speech, which looked to work well. Combined with voice synthesis (which is now pretty good, though still w

I've seen a demo of resyncing lips movement to dubbed speech, which looked to work well. Combined with voice synthesis (which is now pretty good, though still well short of well-acted lines), you could already toss together something for pretty major edits to dialog.