Between a lean proof, and a peer reviewed paper, the former is a lot less likely to be mistaken...

Nothing is perfect.