The system we have on Stack Exchange works great for specific technical questions, e.g. "I want to achieve X, but Y doesn't work because Z. How can I fix Y?". Problems that can be definitively answered in three or four paragraphs and/or a little bit of code fit in well. Generally, there are only a few solutions at most, and sometimes there's just one answer.
The trouble with more open-ended or broad questions is that they can't be definitively answered a lot of the time. It's not necessarily that people haven't tried to get an answer1, but it's quite hard to elaborate on a problem if you don't know what you're looking for at all.
Teaching someone a concept is absolutely fine here, so long as you can fit it within a few paragraphs and provide a reasonable explanation. "How does an engine work?" asks too broad of a concept, but asking about one specific subsection would be fine... even though it's still conceptual.2
A tag wouldn't be something that would help here — Stack Overflow did that many years ago, but it simply encouraged people to ignore the site's standards. Even though you could filter these questions, they're still on the site, lingering, perhaps collecting inappropriate answers or other moderation problems. You can't solve a problem with a group of questions by simply hiding them from people; it just makes it harder to notice there is a problem.
The intended workflow is:
Open-ended questions that can't be answered are closed as too broad or primarily opinion-based as appropriate.
The author takes this feedback into account and edits accordingly (if possible)
The question gets sent automatically for users to review and decide whether it's worth reopening.
The question is reopened (or more feedback is given) and answered.
In theory, at least. In practice, something goes wrong between the question getting closed and getting reopened — as so few questions are reopened. Since 2017, 2341 questions have been closed (most of which have been automatically deleted now), but only 112 questions have been reopened. Unfortunately, very few people are able to turn their closed question into a suitable one for the site, and out of those who do, it's difficult to get your question voted open again3. As goldilocks points out below, some users instead just start afresh and post a new question, bypassing the opportunity for users to review the changes and reopen. In reality, that's necessary if you ever want an answer as reopening is just so slow.
There are some questions that simply can't fit and are inherently open-ended. Users are welcome in chat to discuss these sorts of questions, though, but the issue with that is the chat room isn't too busy and your chances of getting a helpful answer aren't too great.
I think if you're interested in helping people to think about a problem, a back-and-forth chat room discussion might be helpful. Using that, you can work with the OP to understand what they actually need, so that they can then come back and post a more specific question.
1 A statistic from the Stack Exchange team is that ~90% of Stack Overflow users did search before asking. It seems unfair to call someone lazy when the evidence is that they have tried. That said, the search facility here is not the greatest, and if you don't know the term to search for, you could waste days and find nothing. Sometimes I feel like the insistence on 'effort' misses the point of Stack Exchange. There are often problems with open-ended questions, but relying on 'effort' to make a good question is just going to leave you disappointed. A poor question with an attempt that is entirely hopeless is no better than a poor question without... but the one without might help other people in future more.
2 You often hear complaints that conceptual questions haven't put in enough 'research effort', and that users are being unfair by expecting us to answer their questions. That said, my most upvoted answer on Stack Overflow is in response to what could be considered a rather poor question. It could, if you knew where to look, be answered by checking the documentation. When I answered it, the question was actually downvoted to -1. Today, the question stands at 97 upvotes. It's reached over 60,000 people—a perfect example of a question serving many people in future rather than just one today. That would have been 60,000 people not helped, had the user been turned away due to a lack of 'effort'. Future questions can now simply link to the existing one, so there's no need for it to be asked again and again in duplicates. That's how the site is supposed to work. It just happens that a lot of helpful questions aren't considered 'good' questions by our standards.
3 When was the last question reopened by 5 users? A couple of months ago? Turns out it's actually been over 4 years since the last reopening by users, in July 2014. It's only ever happened 8 times in the history of the site! It would seem that there aren't many people voting to reopen (probably because few people even have the privilege).