Wouldn't it be great if this sort of mechanism was enacted for e-mail. Many e-mail spammers have resorted to tactics that are nearly criminal in nature. Equipment and bandwidth exploitation is rampant.
I saw an interesting item recently. There's been talk about 'charging' for e-mail, making it cost to send e-mail to people. I and other don't think that will fly, but there's an alternative floating around whereby when a e-mail server receives an e-mail, it sends back a very hard math problem that must be calculated by the sender's mail server. Then the sending mail server sends the answer to the question and if it's correct, the receiving mail server allows the original mail to come through.
This won't matter much for one-off mailing, but for spammers, this will drive their delivery times through the roof. Think of it: each time a mail goes out they have to spend 1 to 5 minutes of computational time. Right now they can send out thousands of e-mails per second. Under the new system, they wouldn't be able to hit that mark.
Alternatively the computation could be something along the lines of a distributed.net completion block or 2, with a third party like distributed.net getting the computation and verifying the receipt of the block.
Interesting ideas, to be sure... ...Ross...