Wait, "once in a century"? How many times has this been proven such that they feel comfortable talking about the statistical rate of proofs?
(Alas, while “most important” conjectures are a renewable resource, lay reader tolerance for such headlines may not be.)