are these Naive Design Questions?
Before my esteemed fellow moderator Air chimes in, let me start out with stating that "Naive Design Question" isn't a formal reason for closing a question. At best, it's a subcategory of "Too Broad". After some contemplation on the term, I wouldn't mind seeing the term "Naive Design Question" disappearing from our regular lexicon.
With that said, let's dive into some of your other questions / concerns.
Since I happen to design composites, I took some initiative and answered with what I knew. Still, it was even pointed out that the question could be posed better ...
To be honest, I don't see anything wrong with your answer. The question could have been scoped a bit better, and that scoping likely would have made it easier to provide a higher quality answer. But there's certainly nothing wrong with the answer you provided and the OP presumably agreed which is why they selected your answer as answering their question.
It is still a low view / low vote question and answer. I suspect that may be due more to the domain of the question (i.e. niche) as opposed to anything else.
You also cited several examples of other questions that are apparently generating some concern about topicality.
How to reduce Engineering Bill of Material Cost for electronics products?
Was your first example. This question is more about bill of material than about material selection. HDE's comment of "Doesn't this come out to 'How do I reduce the cost of a project?'?" is pretty spot-on with addressing the challenges of this question. Frankly, it's big.
It's a good question in the sense that it's something engineers need to think about. But the size of the question makes it a challenged fit for the StackExchange Q&A format. It currently has six answers, and a few of those are "In addition to the above..." type answers which are hallmarks of big questions.
In any case, I think this is a poor example from the perspective of your question because it's not really about materials selection.
Which material is most efficient for the skin of a solar balloon?
Was your next example. It's definitely within the domain of materials selection, but it certainly could afford some additional scoping.
Many / most SE denizens won't answer a question where it looks like the question is really just a fishing expedition. There's minimal to no prior research; the OP didn't provide any potential solutions; and some of the requirements read as "I want to have it all!"
Going back to the checklist you provided, the OP didn't really address any of the questions you suggested.
I might go so far as to suggest that this example should be closed as "Too Broad" or "Unclear what you're asking" in order to encourage the OP to provide enough detail to sufficiently scope the question.
Looking for best material for a cryogenic ball mill cup and balls
Was your last example. To me, this reads as a good question in search of an expert who can answer it. The OP addresses what they're trying to do; what they have already tried; and possible solutions that they've considered. It seems like the only part missing for this question is a knowledgeable enough expert to answer the question.
To wrap things up:
So, the big Meta question I'm asking is ... what should a good material selection question have? Does this sound like a good start? More importantly, what should it not have?
I think that what you propose is a good start.
As alluded to in the last example, I also like to see a clear statement of what the person is trying to do. I understand that not everything can be revealed due to corporate confidentiality and whatnot. But the more information the community has about what's being attempted, then the better the community can provide an answer or guidance.
Likewise, knowing what the OP has already done or considered is critical information. Why bother suggesting material
ABC when the OP has already tried that and it doesn't meet their requirements? If they don't tell us about what they've done then we end up wasting our time and theirs. And that's also why knowing what else they've considered is crucial. It helps exclude answers the OP doesn't care to hear about and also provides the opportunity for the community to correct mistaken presumptions about potential solutions.