So your argument is basically observational epidemiology study = low quality hence we cannot make conclusions? It would be a fair point if this were a single observational (non experimental) study. However when you look at a dozen studies on humans which all exhibit the same pattern where high PFAS in blood correlates with higher cancer rates then the burden of proof shifts to prove that PFAS is safe, not the other way around.

Theoretically it's conceivable that there is a separate, common factor in all the studies that causes the illness (e.g., people who are prone to cancer are somehow attracted to working with PFAS) but I think after the n-th human observational, plus experimental animal studies that at least show hormone disruption - we probably should on a go forwards basis operate with the assumption that PFAS are likely very toxic for humans, unless proven to be mostly safe.