I use the method where a track gets exported out of the mixing phase and the WAV gets loaded into a new project.
Assume that the mix has volume adjustments
With that said, would it be better to:
(A) - Lower JUST the master channel to a certain amount, then export as one file to mixdown/master
(B) - Leave the master channel volume alone, but lower each individual channel so that the master reads a certain amount of db, then export as one file to mixdown/master
(C) - Don't alter any volume settings when you know the track is "done" being mixed, export each individual channel as a WAV and then recompile everything back together in a new project and proceed to mixdown/master