I wrote a simple perl wrapper for my md5sum differ idea, and it does work well, but it’s slow, mainly due to the fact it’s checking large files across the network. Not much I can do about that right now, but it’s a start…