The problem is it was called a proof beforehand.

So, their is cases of demonstrated lag between publishing and problems showing up. Therefore in the strictest possible terms, every published result is in some way suspects as they simply have yet to be disproved.

I am in no way suggesting this is some kind of reasonable standard. Just that math is slightly more subjective than often portrayed.

PS: This line of reasoning was simply in response to the suggesting that proof somehow inherently implied accuracy.