In that sense the computation no less useless than any other cryptographic technique that has complexity to thwart brute force attacks.
[I may be completely off base here, I wouldn't call myself a bitcoin expert, but this is how I perceived it]
There is another mechanism that could reduce the need for proof-of-work though: the participants could instead show proof-of-stake[1], i.e. prove that they control a certain amount of currency. Somebody has created ppcoin[2] (a bitcoin fork) to experiment with this.
...and if these Bitcoin-mining ASICs are primarily purchased by people trying to attack the system?
Here in the US Midwest, in museums there are old coupons, tokens, wooden nickels printed by local banks, business, bars(!) and pressed into service in the 1800's when coins became scarce. The US Mint didn't respond for years, leaving us essentially without money. Caused a local recession.
Bitcoin may be flawed, but at least its obviously and predictably flawed. This stability should in the long run allow for much better planning on an individual level.
It is fascinating how people may misconstrue the plainest topics to turn them into something completely different. The subject of Fed creating the money has very little to do with current topic, and I did not express any indication that I somehow not worried about it (if you want to know so badly, yes, I am worried, very much, are you happy now?). However both nature of this problem and its causes are completely different from the topic at hand - and I specifically only compared bitcoin to gold and not paper money to avoid any temptation to this completely irrelevant turn. Of course, I was destined to fail.
It just shows the fundamental lack of equality or fairness in the 'currency'.
A company pouring $1M+ to develop a 65nm Bitcoin mining chip: http://blog.zorinaq.com/?e=69