But the point is, and that's why I think systemd is garbage, that there is nothing there that helps me debug what is going on. If the shutdown is hanging there is nothing to do except the reset button or waiting for an eventual reboot. In neither case will there be any forensics about what exactly systemd had tried to do and how it thought that it failed.
That is the part that is not acceptable.
P.S. I shouldn't have to manually adjust a timeout on disabling swapspace on shutdown to prevent a need for a reboot by reset button. The kind of thinking that got big timeouts into those things is precisely why I say that systemd people just think differently than I do.