There are many different kinds of why in philosophical logic. We distinguish at least deductive, inductive and reductive reasoning. Epistemic logic might differentiate more, I don't know.
It should be notable that how and why are reasonably interchangable, as well as in translation wieso, weshalb, warum, but perhaps more commonly wie. Why would you say that, Wie kannst du sowas nur sagen (wie... so, was), how dare you, wie kannst du nur, why would you do that, what did you do that for, Oh Romeo wherefore are you, ... As we can see, the language offers a delicate selection of question words, and many more ways to paraphrase.
TL;DR: If people are asking imprecise questions, how (or why) should anyone know what they mean.
Simple answer: The most trivial questions occure repeatedly. I consider it a matter of didactic and pedagogy to deal with them appropriately, which might be impossible on such an impersonal level. Less frequent questions will be tendentionally even more complicated.
Long answer: If people use the site as extended web search and ask for "Dative, huh?", it should be closed for showing no own research, if a google search would in fact solve their problems. It might be necessary to be defensive here, erring on the side of caution. Less is more.
If a level of understanding is shown that warrants such a question, it might need motivation. Depends on the case. Constructing an example, or analysis of an example should be encouraged. This is so in particular because even seasoned editors might appreciate clever examples, too, but more so because asking the right question is half the way to an answer.
In the second case, the why-question isn't the target either. It merely sets the focus on word order. It does so only indirectly because, I reckon, they are not certain if the example is correct at all. I would not want to restrict the question on that point, if they already trust that it is probably correct. Again one might argue that it lacks research, relative to the difficulty of the construction.
Overall, in both cases I pressume those are duplicates inasmuch as they require thorough introductions to these textbook exame topics.
It should by the way not be a problem that these topics can be tackled with different approaches. That's not a fault of the askee, although it might be helpful if they offer a particular framework to work in. If this is missing, it should not be overinterpreted to go beyond the common curriculum.
So the basic problem is that a why-question is not self-explanatory about the askee's aims, and they can be open ended like why is the sky blue, why do I exist, or not? Open-ended questions are bad. However, lack of specifity also grants the opportunity to interpret the question appropriately. This is especially true when the askee is unaware of the actual problem. The question should ideally display some understanding and respective gaps in the understanding to come to an arbitrary point of interest.
Thus, to answer your questions:
There is no problem with Y-Questions. Yes/no questions might be more direct, and still require elaborate proof on occasion, but people might more efficiently ask multiple choice questions, or how/why when they are not sure what their options are.
If they know what the options are, the problem lies elsewhere .
If you'll excuse the pun, they might want to know how (in)correct their chosen option is, but they'll do the accounting themselves, you will only need to explain the why. That's when the stern schol-masters insistince is needed that that's just so and they will have to learn it, whereby it will be appreciated if authority is proved and not just blankly asserted.
There is unfortunately no general solution.
There is in principle nothing wrong with closing a question, if the close reason is well justified, but the preset options, and the usual comments are often not justified. Arguably, you are interpreting the questuion at the moment that you deem it out of scope for the site. In case of why-questions this often implies that the existential questions would be basically theological or mystical in nature.
The general rules about good answers apply. Even if you would consider an answer speculatively, it is usually permissable if adequately sourced. It is appropriate to addopt a question and edit it appropriately, if you are going to answer it yourself. Closure or deletion are in essence answeres as well.
However, the way you set up the questions with number 4.1. as the default case does not bode well for your general additude to these problems.
Corrolary: There should be no grounds to default to point 1 if only you don't know a suitable source, if you disagree with a particular line of research. Because, a maximally broad scope does not impede on a more narrow scope as you might have it in mind.