[As a quick disclaimer, this is more a thrown-together attempt to explain my approach than to "win" in a disagreement. These different approaches are presumably why it takes more than one person to close a topic. Also, to be honest I most often just plain ignore questions like this one, but it's too late now, obviously]
I tend to take questions as presented. If I don't follow or there's a problem with them and I think they can be salvaged, then I bring it up for adjustment/clarification in the comments. Beyond fixing typos, formatting, and such, I mostly consider the actual content to be the asker's responsibility to adjust, since they're the one asking the question.
If we answer the question we think was asked, it's easy to end up at a completely inappropriate response(a minor recent example), especially if the question was flawed in the first place. That said, there's nothing here about SEO. It's often part of considering which tag to use, but ultimately that was not part of the question and only helps to justify it because you put it there. What search engines do with the tag is their business(and ours to speculate on). It's purpose has more to do with the document outline.
I'm not sure I buy the "webmasters who can't be bothered to read RFC's and W3C working drafts" point, because then what did the asker encounter to find out about the tag in the first place? Also, when you have a standard that isn't even fixed, not being bothered to check the spec isn't really tenable.
As far as coming from Google, I agree with you insofar as topics that require some work to find(yes, that's subjective) or are buried deep in docs with dozens of options and three ways to code everything, eg. Google Analytics. In this case, going back to basic research, a simple search for "html5 section" or even "section element" should have been more than enough to answer before having to resort to SE. I take your point about providing knowledge, but "here's the central documentation that should be your first reference" isn't not providing knowledge. Now, if the asker reads that information and still doesn't get it, he can come back and explain why, and that's a real question. As an exercise, if someone were to ask what the purpose of the p tag is, what action would you take? If different from this tag, why?
The actual question was a fairly flat "What is this thing?" and that's pretty clearly answered in the first sentence of the documentation, no digging involved. As I see it, this is pretty much a canonical example of failing to do research first, even if only to say "Here's what I found and don't understand; please explain it to me." (A valid request for help, versus doing someone's research for them.)
If this had gone even a small step further, like asking why and where to use section versus div versus aside versus article, or even about some particular detail of the tag, I could see it as a real question; note this is also brought up in the spec but starts getting into fiddly stuff not everyone gets easily. But as it stands, this wasn't a request for explanation but a request for a definition and that's what dictionaries(ie. the spec) are for.
