The "Please edit to add the research you have done" part of that close reason is easy to abuse, but it does have a purpose. I think that the "lack of research" attitude could be the start of a problem, so I'm glad you brought it up. However, we should not absolve authors of responsibility for the fate of their questions. If someone ignores guidance from the community, it's not entirely the fault of the close voters when the question doesn't get answered.
We shouldn't ask for research because we want someone to prove that they have made an effort before we will deign to answer their question. The primary reason we need to know what someone has found when they tried to answer the question is so that we can write a good answer. The research the author did shows how they're thinking about the question and what they already know. Another reason is to make it easier for the next person with the same question to find an answer here on ELL if they use a similar research tactic. If the question on ELL contains the search terms that didn't help the author, the next time someone uses those search terms, a helpful answer from ELL might show up in their results.
If the entire question is "What does this mean?" as in example 3, it's difficult to answer well because we don't know where the problem is. The "context" we're asking for is not just the words around the phrase being asked about. Context is also the answer to "Which part is confusing and what do you already know?". It can be difficult for fluent speakers to see the issue with language that seems quite natural and obvious to us. It's like the classic duck/hare optical illusion:
Fluent English speakers might be predisposed to see the "rabbit" in a sentence and non-native speakers might see the "duck". If the question is "what is the mouth in this picture called?", the answers are not going to be very helpful unless we know the person asking the question is seeing a duck. It might also help to know whether they are aware that the picture could also be a rabbit.
Why do we close questions? That's what we need to agree on. In my opinion, we should close questions only when we want to prevent answers to them.We might want a temporary block on answers because the question may change significantly when we edit/clarify it. It might be a permanent block because we want to direct all answers to a similar "duplicate" question so that they can be grouped and ranked. We might want to prevent users from wasting their effort on a question that is completely off-topic for the site. We might want to prevent extremely low quality answers, like copy-and-pasted definitions from a dictionary. If you can't come up with a good reason why someone should be prevented from answering a question, then you should probably just down-vote the question instead of close-vote it.
There will always be some disagreement on whether individual questions should be closed. Also, since almost all of the community is human as far as I know, mistakes will be made. Luckily, it's easy to reopen a question.