I'm coming from the discussions in the comments here.
I disagree that voting up that question (as I too did) implied having a consensus on a homework/self-study tag (as I, for example, did not).
So I would rather have this as an explicit discussion here. Here's just why I think these tags are not useful (but again, up voting this question does not imply - imo - that you agree with me, rather agreeing that this is a relevant issue):
Content wise
We cannot infer quality of questions from these tags. Almost all other "undergrad questions" could have been from their homework. Almost all questions from this guy are (i) copy paragraph from Mankiw that has question and answer (ii) say why his intuition gives him different results (iii) ask for the correct intuition. Does that make them homework questions? No, because it wasn't his homework. Self-study, perhaps. Does it make the content any better/worse compared to a homework question?
Where is the distinction?
All of us are trying to learn Economics. That's the process of dealing with questions that come to mind. Should we add self-study to all the questions now?
What if someone, as a homework exercise, has to answer an intuitive question similar to "Why do we call trade deficits so despite the importers paying the exporters in currency?" He then, trying to grasp the matter, asks this. Now, it is still a "homework question", but differently phrased. Should we tag anything that could be used to do homework?
tldr;
There is no special information in whether sth is homework or not, in order to warrantee such a tag. For these matters, we already are testing academic-graduate. Furthermore, any question can be a homework question (with good or bad effort to hide).