Your question should be asked in a way that is written in the best interests of the reader. That's the most laconic way I can imagine to express it.
That boils down to including all relevant information, clearly stating the problem, giving the question a title that suggests the content of the problem, giving any relevant (minimal) examples, stating what you've tried or searched for, etc - but what it boils down to is just having some empathy for the reader, like anything else you'd write.
So, your question only needs to be useful to someone else that has the problem you're having; it's not your burden to prove that your problem is something that's likely to happen again somewhere else in the universe.
The exception to this is what too localized was trying to capture in spirit - stuff that's essentially an OH DUH! moment (a typo, running 'make' in the wrong directory, not getting enough coffee or sleep, etc) - at which point there's some concern about leaving stuff around that people will likely trip over while searching. That kinda stuff should probably just be deleted once it becomes obvious that there's no point in keeping it.
What you're seeing is a reaction to what some of our most experienced users see as a problem with low-quality questions, and their frustration at us for not doing enough about it. While Robert Harvey is correct, there are no concrete rules of voting, folks used to hold off on down votes quite a bit more judiciously until the floods started coming in.
We're rolling out a wizard for new users to ask better questions (goes into testing in June; I'll have a post about it soon) and tackling what turns out to be more of a guidance problem on our part than a quality problem of the masses. We just didn't do a good enough problem helping folks get off on the right foot, and we made the tough choice of kicking that problem down the road to meet urgent business needs - but we own that, and we're fixing it.
I think, once we have a site that's around 277% less pissed off at what they perceive as folks being lazy (mostly, because we completely fail at guiding new users properly), this will self-correct to a large extent. Then, we can look at what else we might need, including a self-assessment of how relevant our moderation tools are now that we've solved a lot of the problems that we initially set out to solve.
But, (and I love to show this), much more stuff gets accolades than resentment, as evidenced by voting (up / down) trends, from the live firehose (25k+ rep needed):
You don't often see folks coming here and saying THOSE JERKS PROVIDED ME WITH A FANTASTIC EXPERIENCE!!! - but this doesn't reflect the new user experience as well as it should, and it doesn't show our (growing) attrition rate.
But what I really hope folks do is blame us, as in the company, and stay in dialogs with us while we work to fix the experience. The people that just give their time every day deserve your default assumption about them to be good, way more than we deserve it.
But at the same time, any on-topic question that can be understood and answered, even if requires rare knowledge or depicts a problem that wasn't something silly and totally unrelated to what was initially explained deserves a place on the site, an answer, and some appreciation for the time people spent creating a lasting artifact.