I'm not sure it counts as 'failing' when programs are under-funded, then de-funded entirely.

That's the point. A budget that looks underfunded to NASA is looks fantastically lavish in the context of ISRO (the Indian counterpart to NASA).

In a startup failing and iterating is a goal. Why should this be different in a government organisation?

(Assuming you learn from your failures)

If you could before, and can't now, isn't that pretty much a definition of "losing it"?

Two.