[OliJ]

mmmm...
couple of big questions there! I'm not sure from your post wether your doing the mix or not. If you are, then I would suggest that LE is not up to it. Firstly track count will be a problem. Streamlining every thing down to a bare minimum, you may just get away allowing 8 trks for dialogue, 16 Fx, sixteen Atmos, 3 to 6 foley and 2 to 6 music (total track count, ignoring stereo- and this is very conservative, excluding the dilaogues most small budget features would double to triple the track count). You would have to premix dials,fx, etc. down to fewer tracks, and although this is entirely possible, and in fact the way it was all done in the not too distant days of Multi-tracks, you're going to look pretty inefficient when it takes you half an hour to go back and change a pre-mix, then re-import that into you're final mix session, for one of the countless changes that a producer/director are going to ask of you.
Also, Le has no provision for surround panning. Depending on the type of project, this may not be an issue. Further more as far as I'm aware of, there are no RTAS plugin surround encode devices, so you will need a hardware encoder to do your stereo LtRt for the deliveries ( this is all assuming you have some sort of monitoring matrix or surround mixer to assign your LE outputs to the suuround field). Outputing 5.1 stems for archives is also going to reqiure multiple passes, doable
of course but again time consuming. It all becomes a little like " How long is a piece of string"?
Tracklaying is definatley possible, as the 48 tracks now available to LE mean that as long as you make a different session for your dialogues, Fx, Atms etc, track count is not a real issue. You can do premixes of each session and import them into the others as a guide. The lack of suuround panning can be ovecome by duplicating a region accross the appropriatley assigned tracks and the doing the neccesary fades to affect a pan. Of course a TDM system would make it all the easier, but I guess the real question in regards to affordability, is how much more work of this type are you going to do?(though it's a pretty good excuse to splurge out I think).
I hope this helps some, there is a lot more to it of course, but in the interests of brevity I hope I've gotten down to the kernel.

PS. One little hint. sorry if this is old news to you, but it sound like you're relatively new to post production. When laying surround atmospheres, as a rule of thumb, it's best to lay mono tracks in you surrounds (you can duplicate the same track in the Ls as the Rs, as long as you have a long enough delay between them) as when you Dolby encode, any mono component in present phase coherent stereo tracks will end up in the centre speaker. Of course this doesn't go to say you cant do it, but if all your suuround tracks were stereo pairs, the final print would sound very mono as compared to your tracklay.