I often think the same, in fact in my todo list is (in the background) think of more ways to reduce the bureaucracy I face on a daily basis, which takes many forms - from arguing with reviewer #2 over whether or not my finding is notable, to trying to convince a panel that my idea is worth funding.

From a system design perspective, alas resources are limited and have to be allocated somehow, and in the absence of any metrics at all as you suggest, it would ultimately come back to who can tell the better story about their work. This is still at best only weakly correlated with the actual quality of the work.

And so we run up against https://en.wikipedia.org/wiki/Jerry_Pournelle#Pournelle's_ir...

The situation could definitely be better than the current one, but alas I don't have any magic solution.