I find the value of the comments varies a lot by topic. Comments are good in areas where there are at least a few domain experts who frequently visit. This is HN at its best.
Outside of that it's much more of crapshoot. The trouble is that if nobody has any particularly deep knowledge of the subject then it's a case of the blind leading the blind. People will make arguments that sound reasonable and those will gather the most upvotes. Even if they're wrong or nonsensical in the context. It can be worse if people think they know about a subject purely because it's adjacent to (or looks similar to) something they do know about. Then you get strong but uninformed opinions.
In this environment a domain expert can be at best ignored or even downvoted if their comments seem to go against the grain. So they are in the position of either having to do the "do you know who I am?" dance (which some people don't like) or hope someone does it for them.
> The trouble is that if nobody has any particularly deep knowledge of the subject then it's a case of the blind leading the blind
I'm in bio/medtech and anytime there is an article about health, food, neuroscience, or any other 'bio-y' thing, the comments are just as you describe. Sometimes other 'bio-y' people will chime in and set the expectations correctly, but usually there are a lot of comments that are just nonsense (or at least it feels this way to me).
I also feel this way when any article about black holes comes up, however, I am not an expert in astro/particle physics. Still, you tend to see a lot of child-comments from physicists that try to set the parent-comments straight, though the wrong outweighs the correct most of the time.
The Gell-Mann Amnesia Effect[0] should be considered a phenomenon everywhere on the internet where discussion occurs, not just in media, because most people claim far more expertise on subjects than they have. I just assume, now, that anyone on HN discussing any subject other than programming is probably wrong unless they specifically mention that they have direct knowledge in the matter. Otherwise, they're as credible as an anon on 4chan.
In the 90's I kept arguing [to deaf ears] how anonymity is useful to a limited extend. My idea was that we can have real identities but we can anonymize credentials too. You can continue to call yourself Donald Duck and still be a lecturing professor in computer science.
The denial of that possibility continues to this very day on Wikipedia. There it is argued that validating credentials of named professionals is not possible.
In my mind it is as easy as adding a (hidden) token to the html for the professional profile on the university website or nuclear reactor. (Should also have a machine readable tag with the professional title in it)
Say for example:
You add a link to your professional profile to your HN profile, (check a checkbox if you want it to be on public display)
HN then provides you a html sniplet,
you paste it into your uni professor page,
HN crawls it and appends your title to your username.
The funniest anecdote: 2 users on a forum spend weeks disagreeing and fighting over subjects. Their blood was truly boiling. At one point one asked: How old do you think I am? (As a rhetorical) Then it turned out one of them was 14 and the other 72. It completely explained every single disagreement they "enjoyed".
What you describe basically seems like IndieAuth[0]. It would seem to be a good way to validate account identity as well as a way to do so without having secure credentials on the server, for an anonymous-by-default forum.
There's a paradoxical incentive to write comments about topics that nobody knows about. It is true that bad comments get downvoted on subjects where there are many experts. But there might be many of us that for no good reason would like to get access to these Hacker News features that require more karma, and could be tempted to only comments on topics where there are no experts, hoping to reap some karma points doing so.
Outside of that it's much more of crapshoot. The trouble is that if nobody has any particularly deep knowledge of the subject then it's a case of the blind leading the blind. People will make arguments that sound reasonable and those will gather the most upvotes. Even if they're wrong or nonsensical in the context. It can be worse if people think they know about a subject purely because it's adjacent to (or looks similar to) something they do know about. Then you get strong but uninformed opinions.
In this environment a domain expert can be at best ignored or even downvoted if their comments seem to go against the grain. So they are in the position of either having to do the "do you know who I am?" dance (which some people don't like) or hope someone does it for them.