The intention is that the user be able to set up points in each of the files being compared, which are supposed to match up. This helps the code-matching algorithm figure out how the two files relate to each other, in cases where the differences are complex.
I’m still figuring it out by trial-and-error, setting a point in one file and then a matching point in the other. It’s not working as well as I’d have hoped; I’ve had the sync points get into a mismatch state. But it does make things noticeably better.