I have only semi-understood "side-chaining", and never thought I'd need it, but what I now want to do sounds suspiciously like it. Here's the task:
 
I have two tracks which together simulate a slow attack vroom sound like an old bomber engine. They fade in and out, and I added pan effect to simulate a bomber flying across the sky right to left. Then I tried adding reverbs, but the reverb effects seem to pan along with the tracks. This sounds unnatural. What I want is for the actual source audio to pan full right to full left, but the reverb should be independent of that, (and probably mostly remain centered). As it represents reflections off buildings or whatever, it makes sense that the reverb would not pan with the airplane. 
 
I tried creating a bus with the reverb in it, but then I also want to control the volume of each of the two tracks independently, and the volume envelopes don't have any effect because, I'm assuming, the bus I'm sending to is where the volume is now controlled.
 
How do I pan the reverb on these two tracks independently of the actual audio reverb, and also be able to control the levels of each track separately from each other?

jkoseattle
How do I pan the reverb on these two tracks independently of the actual audio reverb, and also be able to control the levels of each track separately from each other?

This has nothing to do with sidechaining, so you needn't go down that path.
 
Because you have two tracks feeding a reverb bus, change the reverb bus interleave from stereo to mono and it will remain centered. Controlling the level of the two tracks separately is easy; use fader automation. The bus level will control the master level of the two tracks, but that doesn't preclude controlling the level of each track independently. Just make sure the track's send to the reverb is post-fader and not pre-fader.
 
 

^^ Also to clarify side-chaning, it is using a separate source to drive the guts of an effect on a track. Ducking is the most common example, where a track's compressor (say rhythm guitars) takes its "threshold" input from a second track (say a vocal). When the vocal rises, the compressor "ducks" the guitars to give the vocal room.

This allows for alleviating frequency masking dynamically where simple mirror EQ might sound awkward or not work (i.e., static solution).