For a second, I felt you were talking about one of my earlier teams that I managed ;)

Excellent insights on failures of teams like this being a problem of a middle layer of management and not just a poor team. In my case, the turn around I brought happened due to a few things:

1) Bring in some principal engineering oversight and invest in ramping up infra.

2) Say NO for not critical features until we got a chance to turn around things (a quarter).

3) Invest in tech excellence and resilience.

4) Make our platforms more open to other teams so that the team doesn't always knock on our doorstep asking for everything to be done.

5) Change the team composition to bring in some strong engineers.