There might be a chicken and an egg problem. If knowledge of theoretical mathematics would be designed like a framework with good documentation what you mention would be solved by simply referring to the corresponding part of the framework.

But to get that more people would need to be involved into transcribing proofs into formal language. Ideally, everyone. This is a higher standard, and somebody needs to ask for it.