I was about to post exactly that. In fact we have a small handful of algorithms where QC *may* be faster in practice, but none of these problems have been shown to be NP-hard/complete. And well i just don't see it really. Factorization takes a huge number of operations on the n-qbit register to factor a nbit number. Nosie etc is not a mear engineering problem. Its a fundamental problem. And even then such a computer can't help at all for a n+1 bit number.