I've found the sequential syncing with mbsync to be a major bottleneck. These days instead of "mbsync -a" I have a script like this, using GNU Parallel: pa

I've found the sequential syncing with mbsync to be a major bottleneck. These days instead of "mbsync -a" I have a script like this, using GNU Parallel: parallel --bar -j0 mbsync "$@" <<EOF mail1 mail2 mail3 EOF