bug 1284933 makes it so systems that run Firefox for less than 120 minutes during a 12 week period are excluded from the orphaning dashboard. This has reduced the number of orphaned users by 45%. It would be helpful to understand how these systems fall into 1. a time length grouping over the 12 week period. 2. number of times Firefox has been run. 3. number of times Firefox has been run for more than x number of minutes. 4. other cases? The obvious thing this could help with is determining whether we should up the priority of using BITS on Windows to download updates when Firefox is not running but there may be other things that can be addressed that this info shows us.

Notes: When analyzing data like this it is usually best to get the first set of base data and let that lead you to more specific data and the items I listed for possible reports are just potential additional reports. This also applies to the interval used to group. For example, when I did the same this for the stub installer cancellations I was surprised that the no matter what interval I used all the way down to 5 seconds the largest grouping was always at the start of the download and had a very long tail.