There are a few questions here I'd consider inter-related, but distinct:
- Are pure "Can I Do--" questions on-topic, even if the answer is invariably "Sure, why not, if you do it well".
- Can we outline a good way to highlight the practical substance of opinion-y questions?
- If we find a good outline in (2), how do we get people to follow it?
As for (2), what the hypothetical outline is, I've grown to recommend a simple formula for Writing.SE questions:
I'm trying to accomplish (X), however, I've run into problem (Y).
This is pretty much foolproof, because by casting this as goal/obstacle, you're pretty much forcing yourself into an answerable, practical Q&A. Only answers that help you reach your goal are helpful, and that becomes very easy to see. Only obstacles that keep you from reaching your goal make sense as questions, and that becomes very easy to see.
So now we get:
- "I'm trying to do (X), but I'm afraid publishers will reject me for that."
- "I'm trying to do (X), but I'm afraid genre readers will hate me for it."
- "I'm trying to do (X), but it's taking so many pages my book gets boring."
- "I'm trying to do (X), but I'm winding up with a really unlikable main character."
Sometimes you'll get "but I want to brainstorm ideas" or "but I'd feel more confident with examples from published works," but those are genuinely harder to get to, because you need to "justify" why the problem you're posing is an actual obstacle.
OK, that's an approach. How do we get people to follow it?
- Close non-compliant questions. This is an option, but it's a heavy hammer. We've had specific periods when we decided to invest in site quality, be more proactive in closing and downvoting. I think that was appropriate then, but I'm not sure that's the way we want to go as our ongoing, general approach.
- Heroic edits. Often gets the best results -- but, it's a lot of work for the regulars. And often, it's almost impossible for an outsider to understand if he's succeeding at identifying the core of the problem.
- Guidance? We've raised this in the path, but I think never carried it out -- a helpful landing page we can point people to, that introduces the basics of what we're looking for in a question, could be a huge help. At the moment, that knowledge is scattered across a bunch of questions on Meta, and finding them is almost impossible -- we've got a lot of "developed lore," stuff that we've gradually understood, but newcomers don't understand and don't even know to look for.
I would definitely invest in a good Writing.SE FAQ/Intro -- I think it'd make a big difference; let us summarize the biggest takeaways somewhere more accessible.
Lastly, Are pure "Can I Do--" questions on-topic?
My inclination is that they are now, and I think taking them off-topic entirely would do more harm than good. That does make it a lot harder for us to say "Well, listen, you should make this one question better," because it looks little different from a bunch of other questions that seem fine.
That being said, I'd consider downvoting some questions for being vague, unhelpful, or purposeless.