This is a question thats come up many times. To put some perspective into the argument have a read at this article:
http://www.bitcoinnotbombs.com/bitcoin-vs-the-nsas-quantum-computer/Please pay particular attention to the section quoting Bruce Schneiers 1996 book, 'Applied Cryptography' and the effects of thermodynamics on the processing of information, coupled with Peter Shor's algoritm for a quantum computer to break public key encryption. It's east to forget about physical laws that don't at first glance seem to have any influence.
The basic message is : SHA256 is safe, period. ECSDA
might be vulnerable
if a viable quantum computer can be built and run reliably but that at the moment would seem to be some years away ay present!
On a side note I've just finished reading an excellent book by Simon Singh, 'The Code Book'. It was written some time ago, long before Bitcoin came into being but it's a very interesting history of codes and codebreaking. Also explains RSA and why super large prme numbers are so sought after.
Sure, SHA256 may be safe, but what about the other algos? Surely, someone will eventually program a miner that can run on a quantum computer using some other algo like scrypt or X11. Do you think these physical limitations you're referring to affect those algos as well?