In my experience the real killer is doing difficult, but low value work.

It's particularly hard at startups, because every "pivot" is just that: a lot of hard work being discarded. You can try to rationalize it like "we learned something" or "at least we're not on that path anymore", but there's still a very real mental cost.