Supposing the paper does describe a more efficient factorisation algorithm, that does not imply that factoring a 800 bit prime (like the author of this article suggests) would be cheap.