I would divide the production process into 3 phases:
1) Recording or tracking: recording your raw tracks. Typically mono sources to single tracks, stereo to two. During the process apply whatever hardware/software trickery to insure that each individual sound source is recorded as well as it can be.
2) Mix down. Take the tracks you have recorded and combine them into a single stereo track (more if doing Dolby 5.1) that is the best overall combination you can get, in accordinance with the client and your own skills & taste. Applying software and hardware trickery as reqiured to get the best overall track you can get.
3) Post Production / Mastering. Take the stereo track from step 2 and do further processing to it. Originally "Mastering" had much to do with getting a sound that would play well off of a vinyl album. Now it often has to do with getting a sound that will sound "good" over FM radio.
Thats my definition, anyone else have a radically diferent one?
With regard to your own situation, while I use Cakewalk I am certain Cubase can "export" to a single stereo file. In Cake I do step 1, then makeing panning adjustments, add effects, and then "mix down" by exporting my finished stereo file.
Most of the time I do not go past Step 2, but I'm starting to experiment using Sound Forge to "master" my mixed down stereo files.
Good monitors help in steps 2 & 3 by accuately reveiling what your instruments are doing, so you can (hopefully) avoid the scenario where your mix sound great on one system and crap on another.