User:SorinaPopescu tripped filter 279 a dozen times, but they were for the same edit, and it wasn't their edit. The edit was actually made by User:Dilloncoulter whose filter log is empty. If you look at the details of one of the log entries, things are way off. The "Changes made in edit" are completely different than the linked diff, and the timestamp doesn't match any edit made by the aforementioned users (see their contribs: 1, 2), or any edit made by anyone to the relevant page. Something is very wrong!
Filter 279, by the way, is tripped when a user makes a series of edits that failed to save to the same page (checking if their username is not in article_recent_contributors), but here both SorinaPopescu and Dilloncoulter's edits did save. Not only that, the filter did not tag the edits as it normally would -- and good that it didn't, because it was a false positive :)
To recount, we have a number of issues:
- Filter 279 is being tripped more times than there were actual actions
- The edit the logs point to was made by a different user, different timestamp, and diff doesn't match the actual edit
- The filter was tripped when no previous edits by the user were disallowed, which is not how the filter is supposed to work
- The edit wasn't tagged as the filter would normally do (probably because of one of the above points)
The first thing that comes to mind is T168736 which I filed back in June. Whether or not the filter was malfunctioning there I'm not sure, but similarly the logs reported wrong values. I have a strong suspicion these issues are related.
I dug deeper into filter 279, I can't find any entries that are as out place as the above. However Special:AbuseLog/19127087 shouldn't have tripped the filter, since that IP had not made previous edit that were disallowed via AbuseFilter. This may or may not be related, but anyway it also appears to be a bug.