It's worth playing with, but in my experience even chatter that images hard left (say) due to combined amplitude and timing differences often isn't all that much quieter in the other channel. So you run the risk of bringing it more front and center, which may not be an improvement even if it is a few db quieter. But maybe one channel really is clean if the chatters ended up in the null, so worth a shot.
Don't think there are any phasing tricks to hep here (unless the chatters were in the rear lobe of one mic, then mixing channles would cancel them out a bit), except insofar as one more thing you could try is converting to a sum (~mid) and difference (~side) and playing with the image width. Don't know how you particular program handles going back and forth between ms and stereo but the brute force way is create a mid channel, create a side channel, then set up a new 4 track file with two copies of the mid as L1 and R1 and then the side and the inverse of the side as L2 and R2 and then playing with the relative levels of pair 1 and pair 2. But in effect this is just adding a bit of L to the R and vice versa, so perhaps you can do this more directly. Or just another 4 track file, but now tracks are L, R, R copy, and L copy, with "copies" lower i level and reversed sides.