Because they won't. The various algorithms that create the "publish or perish" incentives actually reward this. Academic reputation isn't a term in those algorithms, so it isn't rewarded. Same for verifying other's papers and other parts of science and research that aren't actually publishing papers. This is the cause of the reproducibility crisis in academic.

"Show me the incentives and I will show you the outcome"

Reputation is a factor, because most processes involve a subjective evaluation by other experts in the field. And non-experts use service activities such as peer review as a proxy for reputation. If reputable entities use your services as an expert, you are probably a reputable expert in the field.