The question Do we know if the human brain follows Moore's Law? was volunteered for deletion by its author, and I guess one or two other members also think it's worth deleting. But I'm hesitant to actually delete it, for two reasons:
- It's off topic, sure, but other than that there's nothing special about it, as far as I can tell. It's not particularly different from many other off topic questions we have that aren't deleted.
- More importantly, I don't like unilaterally deleting things unless they're actively damaging to the site (spam, severe trolling, sensitive personal information, and so on). By way of justification, we want to have limits on acceptable moderator actions that don't lead reasonable people to worry about censorship, and that's where I think the line falls for deletion.
Don't get me wrong, I'm fine with questions being deleted by the community (3 votes from 20k+ members), and I'm fine with deletions done by the automatic scripts that SE runs. My hesitation only applies to unilateral deletion by moderators. (See also this meta question. This paragraph echoes my answer there.)
Questions which have been closed for 9 days, have zero score, no upvoted answers, and meeting several other criteria are already deleted automatically. But what about other closed questions, which don't meet the criteria for automatic deletion? For example, those that
- have upvoted answers (like the one that prompted this post)
- have upvotes themselves
- are recently posted
- are recently closed
- have reopen votes
As a rule, should we (moderators) be deleting these questions, even when they wouldn't automatically be deleted? Or some subset of them? (e.g. delete all closed questions more than 9 days old regardless of other factors) If so, how can we proactively keep the community's trust that we're not deleting stuff prematurely or inappropriately?
Something to think about: deleting a question that has upvotes and/or upvoted answers, simply because it's closed, is especially tricky because several people expressed a belief that the post is useful - either on its own merits, or because it inspired a useful answer. Now, it is a common occurrence that "upvotes lie" - in other words, we know that people will upvote some things that are bad fits for the site. But I don't think it should be part of the moderators' regular job to make the distinction between "popular and useful" and "popular but bad for the site".