Burden Again Placed on Recipient
This proposal cures spam by again placing the burden upon the recipient. In this case, not only must concurrent mail systems be run to support a .mail domain, but an onerous fee of $2000 and stiff domain registration requirements must be met. This puts relief for unsolicited junk mail out of the reach of many domain holders and seems like a cash grab by domain registrars. Furthermore, it's my belief that $2000 isn't going to buy a whole lot if this simply rides along on the current RFC-821 SMTP standard.
A true solution to spam is only going to come when the burden of identification and its costs are met by the sender. This solution would be a system of credentials either placed at the ISP or upon the domain itself that would allow the recipient to determine the origin of the message. Trust levels could be imparted upon that message and would allow for filtering based upon the tastes of the recipient. This prevents false-positives and troubles seen with rejecting messages with blacklists, shares some of the processor time required for the credentials with the end-client/recipient and eliminates the biggest contributor to spam; the implicit trust in the mail protocol.
While I appreciate the efforts and work Spamhaus; this proposal isn't a real solution. Creating a domain for mail isn't going to fix this problem; only fixing the all-too-trusting mail exchange protocol will cure it.