I am referring to the FAQ entry entitled "Can I stop self-pings?"
http://faq.wordpress.com/2006/04/23/can-i-stop-self-pings/
Following the advice in that FAQ entry does indeed prevent self-pings and trackbacks.
However, when reading posts in a web-based feed aggregator, such as Bloglines, clicking such a link brings one to a non-existent page in the domain of the web aggregator.
Real example:
http://wrathchild.wordpress.com/2006/07/14/has-copyright-gone-too-far-or-are-people-being-too-broad-in-trying-to-apply-it/ has a link to http://wrathchild.wordpress.com/2006/07/13/lemmings-in-your-web-browser/ but, of course, since I want to prevent a ping-back, in my article I only used /2006/07/13/lemmings-in-your-web-browser/ (per the FAQ).
When reading that article in Bloglines, however, and clicking that link, I am taken to http://www.bloglines.com/2006/07/13/lemmings-in-your-web-browser/, which of course does not exist.
I've also seen this behavior on non-Wordpress.com blogs.
I expect tweaking the RSS feed to always include the full URI in tags should fix this problem.