My problem with the "General Reference" justification for closing questions is a bit different from the one raised here by phenry: Too often people vote to close a question on that basis without doing any research into the poster's question themselves.
A case in point occurred yesterday. A poster with a reasonably strong track record at EL&U (1,710 points, 3 silver badges, 7 bronzes) submitted this question:
First use of “bunk”
My question is simple: Which came first, bunk as in "bunk beds" or "to bunk" as in to find a place to sleep? Almost suredly, the definition of one created the definition of the other by association, however I was wondering if anyone knew which came first. Thanks.
Within a few minutes of the question's appearing in the queue, a commenter (who likewise has a fairly strong track record at this site) posted this comment:
This seems like a general-reference question.
...and a few minutes after that, the poster deleted his question.
All in all, no harm was done, I suppose; but in checking a general reference (Merriam-Webster's Eleventh Collegiate Dictionary) during the brief period when the question was open, I ran into something odd: MW said that the word bunk is "prob. short for bunker" despite its being (by MW's reckoning) 81 years older than bunker (if we treat a word's first known occurrence in print as its date of birth). Baffled by this assessment, I consulted a second general reference (my 1985 edition of the OED) and learned that its editors considered bunk to be "Of unknown etymology"—this despite the fact that in the OED's estimation bunker was 67 years older than bunk.
A few minutes of general-reference research had turned up several interesting things: MW's weird theory that bunk came first but was probably derived from the later word bunker; OED's refusal to hazard an opinion about the etymology of bunk; and the two references' opposing views on whether bunk or bunker was the earlier word.
When I went back to the poster's question to note these facts, however, I found that the poster had already deleted the question. Because the MW–OED disagreement had piqued my interest, I went ahead and ran a Google Books search for bunk, bunks, bunker, and bunkers, to see what matches that resource would turn up. Last night, after some hours of investigation, I submitted my own question and answer about the relationship of bunk and bunker, focusing on the issues that I thought were most interesting. But I wouldn't have considered the question at all if not for the original posting.
If the OP of the "First use of 'bunk'" question had simply consulted the Eleventh Collegiate Dictionary, he would have found that the noun bunk dated to 1758, the noun bunker dated to 1839, and the verb bunk dated to 1840. If the OP had simply consulted the OED, he would have found that the noun bunk dated to 1815, the noun bunker dated to 1758, and the verb bunk dated to 1861. Either way, according to both references, he would have learned that the noun bunk preceded the verb bunk.
But it seems to me that if the commenter (whom I don't mean to pick on here) had checked those two general references instead of merely asserting that the question "seems like a general-reference question," he might have been struck by the inconsistency in the dates and in the disputed priority between bunk and bunker. Does anyone else see any irony in criticizing a poster for not consulting general references, without bothering to consult the general references either?
One thing I've learned from consulting multiple references in pursuit of answers to seemingly straightforward questions is that "general references" often disagree on questions of coinage date and etymology—and when such disagreements arise, the notion that consulting a "general reference" will resolve the question goes by the boards.
I recommend that anyone inclined to close a question involving etymology or coinage date as "general reference" take the time to consult at least two independent references first, to confirm that a consensus about those details actually exists. It wouldn't hurt, either, to hold off on closing such questions for, say, 12 hours after they post, to give other researchers time to check their own reference works.
Though I have had occasional success in reopening closed or on-hold questions, it's much harder to reopen a question than to answer one that hasn't been consigned to limbo in the first place.
In any event, I'd love to see a little less haste in condemning questions as "General Reference" and in pressuring posters to withdraw them. One of this site's best features is that it permits users to investigate and spell out the limitations, inconsistencies, and even outright errors that general-reference answers may include.