I composed a response to this question several days ago but didn't post it because it seemed tangential to the question that TrevorD raises. The interesting comments (most recently Araucaria's and Dan Bron's) beneath deadrat's answer have prompted me to try again.
I see two basic weaknesses in the way this site handles closing/deleting questions and answers:
By establishing "lack of research effort" as a close reason, we make a poster's failure to jump through what amounts to a pro forma hoop—one that often adds nothing to the usefulness of the question—a justification for closing a question reflexively, without considering whether the question is interesting in and of itself.
Some close voters vote to close questions that are outside their areas of knowledge and interest, which means that they may be unaware of the deeper ramifications (if any) of those questions.
The first weakness is, I think, what deadrat has in mind when he refers to babies and bathwater. Every interesting question is (relative to the usual murk) a baby. I don't think it helps EL&U overall to reject worthwhile questions on the theory that someday someone will ask the same questions in a way that satisfies our "show prior research" criterion. This criterion may well have arisen as an improvised filter to help us reduce the flood of questions inundating this site. But the filter is not designed to distinguish good questions from bad questions; it's designed to automatically disqualify a large number of incoming questions regardless of their value as questions.
If, post-filter, we were left with boatloads of good questions, perhaps the crudeness of the filter wouldn't be a problem; but that's not what I see. I see a bunch of single-word and phrase requests and a sprinkling of interesting questions. Our site needs to retain all the good questions it gets.
The "show prior research" filter seems to be more effective in blocking some categories of questions than in blocking others. If you ask a question about etymology but you haven't checked Etymology Online's coverage of the word, your question is likely to be challenged as not showing prior research. But if you post a single-word request and do a reasonable job of describing the idea you're trying to find a word for, you're in. After all, what practical research is possible in that case?
So we end up on the one hand blocking at least some questions that would have yielded more- complex and more-useful answers than Etymonline has the space, time, and inclination to provide, and on the other okaying many questions that amount to miniature instant trivia contests. Because our "show prior research" requirement is far less effective as an automatic blocker in areas such as single-word requests and phrase requests than in areas such as etymology, it may contribute to the disproportionate number of SLRs and PRs among open questions on this site.
The second weakness mentioned above reflects a presumptuousness (or at least a degree of overconfidence) on the part of some close voters. Whether because they dislike certain classes of questions or because they feel duty-bound to pass judgment on every question in the Review queu—including ones that they aren't especially knowledgeable about—they vote outside their areas of interest and expertise.
I'm well aware of the temptation to pontificate on topics of grammar and usage that one is unqualified to address. There are large areas of linguistics and grammar that I should leave alone, and yet I sometimes get carried away and post an answer to a question in one of those areas—only to discover that my answer is completely wrong. I can tell myself, "No, Sven: Do not try to answer questions about parts of speech," but consistently exercising appropriate self-discipline is hard.
In their zeal to close-vote, it seems to me, some site participants must be permitting themselves to pass judgment on topics that they would do better to recuse themselves from. I hate the idea that we may be discouraging extremely knowledgeable contributors from participating on EL&U because we too often close interesting questions that we don't recognize as interesting.
Here is how I would like to see EL&U deal with the two weaknesses enumerated above:
Stop using the "no prior research" reason for closing, and instead use the "I am voting to close this question because ..." reason, completing it with a brief explanation of why the question itself is bad. This will accomplish three things: (a) it will help us escape our overdependence on a close reason that doesn't distinguish between bad questions and good questions; (b) it will force us to state clearly what we think is bad about a bad question; and (c) it will give other voters a basis for voting to close on the merits (or demerits) of the question as a question rather than on what often amounts to a technicality.
Vote to close a question only if you are qualified to pass judgment on questions of the type involved. And by "qualified," I mean possessed of sufficient expertise in that area to be a good judge of the question's intrinsic merits.
Junk questions are junk, and we should close them without equivocating, apologizing, or second-guessing ourselves. But what makes them junk isn't that they fail to include evidence of prior research; it's that they have no depth and complexity, or that they can have no practical value to anyone but the poster. Let's close them for the right reason, and let's not be so eager to close questions that fall outside our personal areas of competence and interest.