Another site I had not previously heard of.
The failure of upcoming is two-fold. First, they have (practically) no content. Second, they cannot succeed as an independent site.
Remember why email originally took off (LONG ago) on Unix, and took much longer on other platforms? The reason was very simple. Even if you had never used email before, if anyone sent you email you would see a “You have mail” message the next time you logged in. Every Unix user had an email inbox (in the standard setup) so the step from not using to using email was very small.
In the early 1980’s I was trying trying to convince Burroughs folks of the value of email. Burroughs did have an in-house email application (AMS?). You had to run the email application to check your email, and the application was slow to startup and painful to use. Since in the beginning no one was using the email application, there was really no point in firing up the application every day – just to find there was nothing there. A classic chicken and egg sort of problem. I argued that there had to be an inexpensive and automatic check for email every time you logged in, and for every user. I failed to convince anyone as no one had (at that time) made enough use of email to see the value.
Ironically, at Burroughs we spent a deal of time trying to communicate. Phone tag, written memos, hunting through offices - all were popular pass-times of necessity.
Upcoming.com has the same problem. No point checking the site if there is no content. The RSS feed solves the problem of how to check without making a visit to their site, however they still need content and some sort of automatic hook. Very few are going to monitor an RSS feed from a site that has no content.
Jon is right - they badly need an API to add and read events from Upcoming.
Given an API and a means of tagging events, any organization could find Upcoming a useful place to record and (eventually) advertise their events. Without an API and with only manual entry … well, the current level of participation is elegant proof.