the sticky threads here: http://www.gearslutz.com/board/post-production-forum/
have a lot of real life info from professionals.
No matter what format you mix in you can always just export each channel as a mono file and change the configuration when doing the final layback.
When I mix, I always go to the highest playback standard, in this case it would be cinema. When you output your deliverables, most encoders have you drag your individual mono files to the correct channel/speaker and then do the proper encoding. That is how the minnetonka and Apple encoders work and even hardware encoders you can crosspatch to get the channel alignment correct.
The most important thing you can do is to mix in a calibrated room of the correct size at the correct level. Doing that will help ensure your mix translates to a variety of environments.