In general, I think that Gates is being overly optimistic. Some of these ideas are promising, but most of the anti-spam ideas are more likely to change the tactics of spammers than reduce the overall rate of spam. What's interesting to me is that his optimism comes largely from ignoring the problem of insecure computers on the Internet, primarily insecure Windows computers on the Internet.
Schneier notes that would-be spammers can defeat most spam-stopping plans by taking over new e-mail accounts (most of them on computers using the insecure Windows operating system) and using them to send enough new spam to overwhelm anti-spam measures. [Netlawblog]