It seems to me a lot of extra complexity is introduced by the way they use Signed-off-by lines. Theyre trying to use these to record who accepted the pull request into the mainline. Is this even necessary?
How do the Linux kernel folks handle this? (Obvious question I ask every time somebody claims to have a problem using Git.) It seems to me they dont bother with this information, they only record the contributor of the patch, not who accepted it.