Not a major bug, it would have been major if it applied.

Maybe not major, it's important as we are loosing correction edits.
It's already a pain to go back and check if everything has passed for not having such a bug that even prevents you from checking that your edits have not vanished.

The semantic difference between the text "Failed Vote" and "No votes" is a little weak to begin with. In either case the edit failed to get sufficient yes vs no votes in order to pass; it's just that the rules are different. But you're right that there's a valid use case for searching for these in particular; especially given how low our current voting rates are.

Voice, it is not just a light glitch, it is a regression by the way. The failing edits due to high quality data with no votes had their own status, even in NGS (cf. my no votes search link in ticket desc.).

I know they had their own status and I know its a regression. I've been around MB quite a while, remember? This is why I didn't change to "Improvement" and agreed that searching for these was a valid use case.

Regressions are regrettable and thus extra effort should be made to avoid them; but they're not fundamentally any more or less important than any other bug once they've happened. They need to be prioritised independently. Fixing newly created ISEs in Release Editor > fixing this regression; for example.

Agreed to fix within 3 months by http://scheduling.ocharles.org.uk and http://chatlogs.musicbrainz.org/musicbrainz-devel/2012/2012-03/2012-03-26.html#T19-36-54-3898

In review at http://codereview.musicbrainz.org/r/1826/

Oliver, would it be possible to include a modbot edit note so that it wouldn't silently fail ?

Possibly, but that's for another ticket.