The way I normally start when mixing music for 5.1 (mostly in feature films, but sometimes also in documentaries), is to pull the music slightly towards the center, so that it isn't strictly L-R but more L-R with a little bit of center, so you don't get too much of the "hole in the middle" syndrome. Sometimes I pull it back a little too.
Normally I then apply a hall reverb to the surround speakers, leaving it pretty dry in the front and almost totally wet in the surrounds. This way the music surrounds you without having too much of the direct sound coming from behind you.
I will also send some of the music to the LFE channel.
If I have stereo stems of the music available, I will pan the softer elements more to the back than the percussive elements, and be more selective about which tracks I feed to the LFE channel. If there's no particular bass information in the channel, there is no need to send it to the sub.
The more tracks I have available for mixing, the more specifically I will pan them across the LCR front, for instance narrowing the bass down to the center channel (and LFE of course), or pinpointing the close mics of the orchestra across the screen. I may also use an extra reverb in the front, to give the narrower sounds a bit of spread. I will also narrow the reverb a bit so it isn't strictly LR.
The real advantage when mixing in 5.1 compared to stereo is in my opinion more the enriched clarity and higher fidelity that comes from using more speakers, than it is the fact that you can have sound coming from behind.