I've read the other answers here, and the comments.
The two authors (Floris & Alemi) of the (excellent) highly voted answers to the original Coin Question have replied here, and I think that they have useful insights. It's interesting that their answers on this meta question are the lowest rated at the time of writing.
Horses for courses
My suspicion is that one person's good question is not another's good question.
For some, a nicely clear unambiguous question is a good one. Ideally one showing the asker's research and a statement of why they couldn't proceed, or their uncertainty.
As a professional software engineer, primarily using Stack Overflow rather than Phy.S.E., these are certainly the kinds of questions that I like to answer and hope to ask. Stack Overflow's How do I ask a good question? covers this.
Regular users are likely to prefer this kind of question. Those working in the field, and the enthusiasts with an interest in Physics that hang about here and get in the way, like me!
However, for others, a simple interesting question with excellent answers that teaches them something, is a good question. Answers that encourage them to feel science is something that they can do, participate in, use, and understand.
Floris's answer mentions a link on Reddit, and I would think this is the kind of question that would appeal to someone happening on the site from a link and having a kind of "popular science" interest.
I suggest that a well written question, in a case like this, might not be as successful as the quick carelessly written one. As a populist pedagogic tool, it might be less likely to pull the masses in, because an outsider, reading a well written question, won't feel that is something that they, personally, would ask.
Is it a problem?
In general, I think that the first kind of question is the better kind, but I think the second kind has its place and use, as it helps provide a fascination with science.
Others may feel differently if they would rather keep Phy.S.E. as a fairly highbrow tool for learning and answering professional queries – I'm not judging that as a bad thing: having relatively uninformed people about can be annoying, particularly if they reduce the signal to noise ratio.
Is this something to worry about? There I'm not sure. Voting is useful because it provide a mechanism to resolve disputes like this where people have different needs and desires.
I wonder if the experienced physicist and teachers here might have a good sense of when a "dumb" question can be a good tool, and so, in a way, becomes a good question. You probably also need a sense of whether it will have popular appeal, which is hard to know.
In summary, I would say that if you want Phy.S.E to be a place that reaches out to the popular science crowd and gives them an insight to what it is to experiment and use science, then support what Alemi and Floris say. Edit questions to improve grammar, spelling and general readability. Go easy on questions that haven't put in the ground work but sound interesting.