If you use the "review each commit separately" style then I just made some minor but useful improvements to this workflow. First, the file matrix header now indicates which revisions/commits have been rebased away by crossing them out, and simplifies the rebasing arcs so they're always readable (if potentially less accurate):
Second, the algorithm that picks which diffs to review next now focuses only on current revisions and makes better use of the rebasing relationships by following them transitively. So whereas before you'd always be reviewing every commit, now you'll automatically skip over any that have already been rebased away.
There's almost certainly new corner cases in the updated algorithm, so please open issues if you spot anything wonky. Thanks!