This seems like a rip-off of Scott Aaronson's news piece: www.scottaaronson.com/writings/phcollapse.pdf
Interesting, but the article completely washes over the fact that while a polynomial algorithm exists for an NP algorithm, if P = NP, it has not yet been determined; so, while prime factorization would be P in P = NP, we would have to discover that particular P = NP algorithm. Technically, I suppose we could map prime factorization to 3-SAT and use the O(n^$some_ridiculous_number) algorithm to get prime factors, but in a strong majority of cases, it would probably still be more efficient to use the a^N algorithm.