This is a math question. It's an interesting math question — so interesting that it is a research problem whose solution is still unknown.
And the top-scoring answer as I write (note: it's since lost its top position) is a “thinking outside the box” solution, i.e. one that ignores one of the assumptions in the question to provide a “clever” answer.
Having a wrong (yes, it's wrong: it contradicts the assumptions in the question) answer is pretty damning for the image of the site.
Another less spectacular case: 8 upvotes, top score as I write for a nonsensical answer to a question that requires a good understanding of set theory to even model the problem correctly, and an understanding of the axiom of choice to solve.
Here's an even more egregious example on a pretty easy geometry question: want to join points with a line? Declare that a line is as thick as you like! (How many dimensions does a line have if you call it fat?) That nonsense answer got 63 upvotes. These aren't isolated cases: “clever” wrong answers to math questions often get a lot of upvotes. We aren't good at rating answers to math questions.
Given that the bulk of questions on this site are either about math or are what-am-I-thinking-of riddles, that's about half of our questions that aren't getting good collections of answers. Meanwhile, there is a thriving Stack Exchange site for math questions.
Should we declare that if a question involves non-elementary mathematics, then it isn't a puzzle anymore, it's a math question, and thus off-topic?
(Previous, related discussion: Are probability exercises on topic here?)