"I use ProTools LE and sometimes I have to process a track or submix to free up plug-ins prior to the final mix. Is there any advantage or disadvantage to processing with AudioSuite vs. bouncing to disk?"

Any advantage? This question has two components: the sonic ramifications and the functional ramifications.

Sonically, there is no inherent difference between the two methods. However the things you can do are significantly different. For example, let's say you have a track that needs both EQ and compression. Processing with separate AudioSuite plug-ins will require two passes, one for each processor. If you do a mix to disk you can capture the cumulative results of them (in addition to other things you may be doing) all at once. The limit of what you can capture on any mix to disk pass is the same as the overall limit of the system. So you could process a whole drum mix with gates, compressors, and reverbs and distill it all down to one, complete stereo drum mix. But you need to be careful. If you have already written panning and level automation for that track you should turn it off and just set the track level at unity gain. Of course you also need to remember to mute or disable all the other tracks, etc.

On the other hand, when working with small (short) files such as those often used in sound effects and post production work, it can be faster to just process them directly in the plug-in because setting up a mix to disk is at least several more steps by the time you mute the other tracks and get the level set.

Another trick is to do two at once. If you are working with two mono tracks you can easily set up a mix to disk where one is panned left and the other to the right, each with their own processing. The result will be a left and right file, but actually each one of those is one of your original two tracks with its processing added. Clear as mud?