The statistics page for Greasemonkey is visible to everyone. The raw data is even available for download. But it can be hard to read, due to the level of detail and formatting that is applied. So, I've taken the time to analyze it carefully. The first interesting thing that we can see is the usage trends over time:
(Looks like Mozilla had a reporting issue around May of 2009.)
I've also made a pie-graph of app usage, for the average values of the past 4 weeks:
That pie chart represents these numbers:
So, let's say first off: we know this is a bad measurement. There's (almost) no "other" because there's no official support for other platforms, so only third party alterations make this usage possible. Thus, this data doesn't help us answer (i.e.) "Should we support Flock?" or "Should we support SongBird?".
It does let us know a little bit about what versions of Firefox we should support. All of 1.0 and 1.5 make up only 0.23% of the user base. Firefox 3.0 and 3.5 make up 95.44% of the user base. Firefox 2, however, makes up 4.31% of the user base. That's a much harder call.
Hackers: The raw data and charts behind this post is available on Google Docs. You can also see the python script that turned AMO's raw data into this presentation.