I admit it's possible the project "succeeds" and gets out the door. My prediction in that case is that we limp along and eventually give up after maintaining it for a while.
I only work my ~40 hours a week. When I did much more than that, I think my output went negative.
I think I learn a lot, at least. The project has a more "modern" stack than the company's other main product. And management/leadership being hands-off means I make a lot of big decisions myself. Many bad decisions, but at least I learn what not to do next time.
We are reasonably consistent with estimates, but there's this hidden assumption that 1 point equals 1 developer day. So even though we consistently get 20-25 points done per sprint, we typically cram more items to meet that 30 point threshold.
Oh, and of course you may end up dragging items sprint to sprint if they don't get finished.