There's a third possibility: there's a flaw in the proof, but one that is hard to find, and the discovery of the flaw actually advances the field in interesting ways.

If I were a betting man that's actually where I'd put my money because the paper passes by bogometer test (but I am nowhere near qualified to assess whether it's actually correct).