|Perl Monk, Perl Meditation|
In a nutshell, I don't believe we can. Google is as useful as it is because of that caching facility.
As for it violating copyright law, I think that was already decided that it didn't, as otherwise you'd find all companies that operated a webcache violated that copyright.
Personally, I don't have any issues at all with cached copies of stuff appearing on google (interestingly enough, it seems to have archived stuff from the chatbox too.. I found an old comment of mine appearing on there on a search for "Malkavian perl monk")..
Bear in mind, it only appears if someone's actively looking for that particular data.
If someone is actually interested in looking for Malkavian the Perl Monk, I'm pretty sure they'd come to the Perl Monks site, and look at my home node.
If someone's interested enough in me to go and look me up on a search engine, I think I'd count myself flattered. :)
The idea behind Google's web cache is that quite frequently, some very interesting stuff is lost, either by user account shutdown at universities, or machines temporarily falling off the internet. In those cases, google offer a snapshot of this data for a time (won't be forever, as they don't have infinite space to store it), under the understanding it's a cache, and thus likely out of date, but still allows you a glimpse of what otherwise wouldn't be reachable.
Anyhow, that's just my take on it.