My company has been using FogBugz for a while now and we are generally happy with it as a bug-tracking tool. I've been reading Joel Spolsky's articles about their Evidence Based Scheduling feature. It sounds great in theory, but I haven't seen much discussion about how well it actually works in practice. Before I spend a lot of time and effort trying to convince my co-workers to buy in to using it, I'd like to hear from people who have been using this feature in their development.
Have you been using FogBugz' EBS? If so, are you happy with it? Have its estimates been accurate enough to be helpful? With the benefit of hindsight, do you think it was worth the effort to set it up and input all of the information/estimates it requires? Is there some other mechanism that you found that works better?
(Note: I've deliberately posted this to stackoverflow.com rather than fogbugz.stackexchange.com, since I suspect that the user base at fogbugz.stackexchange.com might be unduly biased in favor of FogBugz -- in particular, ex-Fogbugz users who've moved on to something better are unlikely to read or post there)