There are small communities on reddit that act somewhat like the old-school forums (and presumably a small handful on hacker news, although its a bit harder to happen on-site here due to the lack of the channel mechanism, so if it DOES happen here, its more liable to be because of some connection to a real-world/pre-existing network or a specialised subject-matter selectively clicking on stories and augmenting their communication within the greater online community of HN).
My point is, among the wider HN community, unless your friends fall into the handful of 'celebrity' accounts, those names that you know would likely be anonymous names to most commenters who identify as/are identified as HN posters.
The point is ALSO though, that you're able to do this while maintaining the HN identity, and the HN forum interface, even though relative to HN accounts the accounts of your friends are tangentially approaching 0% of users. And you do this because the news-aggregator mechanisms allow you to self-select into the posts relevant to your interests with very little effort, so you don't FEEL like you're excluding from the community/or being excluded against, even though you're not reading most material and they're not reading yours, and most have no idea who anyone else on the platform actually is.
Scaling like this isn't realistically an option for "Cyberspace security, compiler and UI toolkits forum: your number one place for security, compiler and UI toolkits discussions!".
The last points that i didnt mention was also opportunity cost (in terms of how many networks individuals can actively participate in) and network effects (where do most/key individuals end up spending their time resources). The popularity-based metrics of the news aggregator sites are extremely powerful draws of the majority eyeballs relative to the webforums, and then network effects emit another draw even if there are holdout users who prefer web-forums.
By a quick estimate I “know” a couple hundred people on here: their usernames, some trivia about where they work and what their interests are, and what their opinions have been on previous topics; this subset leans heavily towards regular commenters and people who I interact with. On random topics this usually means that maybe 10% of the comments I see are from this group, but on specialized topics (security, compilers, UI toolkits) more than half of the comments I see are from them.