Thankfully his responses are saved for posterity in the imap-protocol discussion list archive. ;)
https://mailman2.u.washington.edu/mailman/mmsearch/imap-prot...
In order to create a new protocol/platform, you need to build an outstanding product. Apple is probably the best at doing this. The Facebook platform is another example, where awesome photo sharing led to the full social network, and later Open Graph.
Microsoft actually made many improvements to a mail/calendar/task sync protocol with Exchange (seriously, go read the spec!) and they had adoption via Outlook, but they made the mistake of requiring licenses for development which kept it locked-down for years.
This stuff is really tricky and just takes a while to happen. I think the email ecosystem is actually poised for a protocol shift, due to the easy distribution+upgrade path of App Stores and the web. But working with mail is a huge technical challenge-- just to start, a single user has more mail data than most startups' entire database.
I've actually been working on a new email app for the last year, so I'm intimately familiar with the difficulties. :) You can sign up for our beta here: www.inboxapp.com