Yes it gives us both a super-predictable queued operations model for using emacs, and serial operation in a parallel world. Years ago when single core performance was king this wasn't an issue, but now we're stuck having to optimize the long hot path.