Stats - imperfect, but with reasonable disclaimers - and quite interesting w/r
the share of detected Exim vs other MTA.
http://www.oreillynet.com/pub/a/sysadmin/2007/01/05/fingerprinting-mail-servers.html
No interest in arguing the merit/lack therof - the information is what they were
able to do 'on the cheap', and the methodology is therein described.
But has anyone seen any other recent studies (flawed or otherwise)?
Bill