(Expanding on this comment)

The key thing missing from your account of my views is that while I certainly think that “local validity checking” is important, I also—and, perhaps, more importantly—think that the interactions in question are not only fine, but good, in a “relational” sense.

So, for example, it’s not just that a comment that just says “What are some examples of this?” doesn’t, by itself, break any rules or norms, and is “locally valid”. It’s that it’s a positive contribution to the discussion, which is aimed at (a) helping a post author to get the greatest use out of his post and the process and experience of posting it, and (b) helping the commentariat get the greatest use out of the author’s post. (Of course, (b) is more important than (a)—but they are both important!)

Some points that follow from this, or depend on this:

First, such contributions should be socially rewarded to the degree that they are necessary. By “necessary”, here, I mean that if it is the case that some particular sort of criticism or some particular sort of question is good (i.e., it contributes substantially to how much use can be gotten out of a post), but usually nobody asks that sort of question or makes that sort of criticism, then anyone who does do that, should be seen as making not only a good but a very important contribution. (And it’s a bad sign when this sort of thing is common—it means that at least some sorts of important criticisms, or some sorts of important questions, are not asked nearly often enough!)

Meanwhile, asking a sort of question or making a sort of criticism which is equally good but is usually or often made, such that it is fairly predictable and authors can, with decent probability, expect to get it, then such a question or criticism is still good and praiseworthy, but not individually as important (though of course still virtuous!).

In the limit, an author will know that if they don’t address something in their post, somebody will ask about it, or comment on it. (And note that it’s not always necessary, in such a case, to anticipate a criticism or question in your post, even if you expect it will be made! You can leave it to the comments, being ready to respond to it if it’s brought up—or proactively bringing it up yourself, filling the role of your own devil’s advocate.)

In other words—

I mean, if 10 years from now, when you are doing something quick and dirty, you suddenly visualize that I am looking over your shoulders and say to yourself “Dijkstra would not have liked this”, well, that would be enough immortality for me.

And this is a good thing. If you posit some abstraction in your post, you should think “they’re gonna ask me for examples in the comments”. (It’s a bad sign, again, if what you actually think is “Said Achmiz is gonna ask me for examples in the comments”!) And this should make you think about whether you have examples; and what those examples demonstrate; or, if you don’t have any, what that means; etc.

And the same goes for many other sorts of questions one could ask, or criticisms one could make.

(Relatedly: I, too, want to “build up a context in which people can hold each other accountable”. But what exactly do you think that looks like?)

Second, it is no demerit to a post author, if one commenter asks a question, and another commenter answers it, without the OP’s involvement (or perhaps with merely a quick note saying “endorsed!”). Indeed it’s no demerit to an author, even, if questions are asked, or criticisms made, in the comments, to which the OP has no good answer, but which are answered satisfactorily by others, such that the end result is that knowledge and understanding are constructed by a collective effort that results in even the author of the post, himself, learning something new!

This, by the way, is related to the reasons why I find the “authors can ban people from their posts” thing so frustrating and so thoroughly counterproductive. If I write a comment under someone’s post, about someone’s post, certainly there’s an obvious sense in which it’s addressed to the author of the post—but it’s not just addressed to them! If I wanted to talk to someone one-on-one, I could send a private message… but unless I make a point of noting that I’m soliciting the OP’s response in particular (and even then, what’s to stop anyone else from answering anyway?), or ask for something that only the OP would know… comments / questions are best seen as “put to the whole table”, so to speak. Yes, if the post author has an answer they think is appropriate to provide, they can, and should, do that. But so can and should anyone else!

It’s no surprise that, as others have noted, the comments section of a post is, not infrequently, at least as useful as the post itself. And that is fine! It’s no indictment of a post’s author, when that turns out to be the case!

The upshot of this point and the previous one is that in (what I take to be) a healthy discussion environment, when someone writes a comment under your post that just says, for instance, “What are some examples of this?”, there is no good reason why that should contribute to any “relational” difficulties. It is the sort of thing that helps to make posts useful, not just to the commentariat as a whole but also to those posts’ authors; and the site is better if people regularly make such comments, ask such questions, pose such criticisms.

And, thus: third, if someone finds that they react to such engagement as if it were some sort of attack, annoyance, problem, etc., that is a bug, and one which they should want to fix. Reacting to a good thing as if it were a bad thing is, quite simply, a mistake.

Note, again, that the question isn’t whether some particular comment is “locally valid” in an “atomic” sense while being problematic in a “relational” sense. The question, rather, is whether the comment is simply good (in a “relational” sense or in any other sense), but is being mistakenly reacted to as though it were bad.