In the past myself/philor have stumbled across failure modes that are not recognised by OF (eg a starring something for weeks on TBPL and OF hasn't heard of it). I believe many of these to now be fixed as of bug 807114 and friends, since we no longer strictly limit OF to bugs containing [orange]. One of the arguments against doing away with the TBPLbot bugspam is that OF "often misses failures". I'd like to either prove this is no longer the case, or if it is, then make sure we fix it - so people can no longer use that argument (and also the bonus that OF is then more useful for me when prioritising bugs to chase). I'm going to export the last couple of weeks worth of change bugs from bugzilla and compare a scrape of the bug list in OF for the same timespan. Obviously, OF will be missing some bugs due to trees=All not covering Try etc; but it should be a good start.

I can't find any more cases that aren't already covered by the dependant bugs, so closing this out.