Reminds me of this post I saw on CompSci Reddit once. A person claimed to have made an O(n) algorithm for finding loops in a graph (or something like that), and he posted the source code. The comments were all in the line "big if true, but I'm going to have to see a peer reviewed paper" to which he answered along the lines of: "I'm paid to write code, I'm not paid to write papers. I made this algorithm solve a problem at work, take it or leave it".

The thread died without anyone actually checking if his algorithm worked or not.

Who knows...