Conventional wisdom holds that you can't *force* users to explain themselves, and trying to force it will just result in users typing "asdfasdfasdf" to fill a character limit (or otherwise typing nonsense real words to bypass an "arms race" filter that tries to block them from writing nonsense).

 > My understanding is that when a question is downvoted, the downvoter should provide a comment explaining why they downvoted and same with close votes (unless it is obvious). I follow this practice and always explain my negative votes unless it is patently obvious what the problem is.

It's perhaps a *convention* or a *good idea*, but it's far from *required* (and users who don't do so shouldn't be punished).

 > Nevertheless, on this SE in particular (I participate in about 10 different SEs) I notice that both I and other people frequently get downvoted or closed with no explanation. A typical example is my question earlier today (Can I run Linux on an Android table?)

...So? The closing action *must* provide an explanation of why the question was closed; the downvotes don't require such action. If your question gets closed, it's safe to assume that the downvotes have been cast for the same reason(s) it was closed.

I fail to see the issue you are having with this.

 > (I also note in passing that on the first two pages of SU, there are 30 questions. Of these 30 questions, only a single one has an upvote, and one has a single downvote and the others all say zero. I don't want to start getting too meta on this, but I think to only award a single upvote to 30 questions, something is wrong here.)

 - That's a very small sample size, and not a random sample; the front page most often has the *newest* questions -- why would you expect brand new questions to have upvotes at all, let alone several? No one's had a chance to look at them yet, except for the ones that get bumped to the frontpage because they've been edited or answered recently. For a better sampling, use data.stackexchange.com and look at questions that are at least a couple weeks old in terms of the time the question has *existed*.
 - A very significant minority (or perhaps even the majority) of questions asked on Super User are, in fact, *bad*. Less-experienced users of the site tend not to invest the time into their question that would be necessary to make it useful for other users after they solve their particular problem, by fleshing out the question with details of their environment; things they've tried; exact error messages; symptoms and the reasons they think those symptoms lead to particular diagnoses; and so on. They'd rather just ask in a casual forum style, "it's broke, halp!" -- expect these questions to get eventually closed/deleted. 
- It wouldn't surprise me at all if any arbitrary 30-question sampling of the front page of new/edited questions is mostly or completely crap. Good questions can be rather hard to come by sometimes.


For fun, I just accessed http://superuser.com and got the following results:

 - Cumulative total score of all questions on front page, subtracting negative votes and adding positive votes (for 48 questions): 130
 - 25 questions had a vote score of 0; of those, 80% were asked today
 - Only two questions had a negative vote score
 - 16 questions had a score of 2 or higher

So as you can see, it really depends on exactly when you access the site, as to which questions you will see on the front page and how good they'll be.