> You can just look at pure user counts and see they are more usable/accessible overall

User counts are a poor comparison when you consider how much the pool of potential users has grown.

As points to reference, Microsoft Windows 95 sold 7 million copies in its first 5 weeks so call it ~73 million copies sold the first year.[0]

20 years later in 2015, Samsung, Apple, and Huawei combined sold ~73 million phones across 5 models.[1] Windows 10, released that same year, had Microsoft shooting for it installed on 1 billion devices within 3 years.[2]

2005 estimates 1 billion (or 16% of the global population) online. 2020 estimates 4.9 billion (or 63% of the global population) online.[3] A million users used to be a big deal not that long ago.[4]

Not that I disagree we've gotten better at some things but Outlook Express (newsgroups) and mIRC (irc) seemed plenty accessible to millions of non-technical users.

[0]: https://news.microsoft.com/announcement/launch-of-windows-95... - when searching for a number I saw first-year estimates at ~40 million

[1]: https://en.wikipedia.org/wiki/List_of_best-selling_mobile_ph...

[2]: They made it in 5, https://www.theverge.com/2020/3/16/21116762/microsoft-window...

[3]: https://www.itu.int/itu-d/reports/statistics/2021/11/15/inte...

[4]: https://en.wikipedia.org/wiki/The_Million_Dollar_Homepage ; https://www.youtube.com/watch?v=4e0n7vTLz1U

Improvement in user counts? I'll give you that. Improvement in the quality of discussion? Not so much. It isn't a coincindence that we're discussing this here and not somewhere on discord.

Remember the eternal september.

One important Usenet feature that enables efficient tracking of and participation in discussion is that read/unread status is tracked per posting. (See my related comment in the sibling subthread.) That kind-of requires keyboard navigation to make it practical (e.g. press Tab to jump to the next unread posting, which then automatically becomes read). But keyboard navigation by itself also tends to maximize usability for frequent/regular users, which I'd argue are the important demographics for discussion forums.

One issue is with mobile UI, where there's really no good way to make per-posting read/unread status practical with just touch (and/or voice) input. But I'd also argue that asynchronous discussions (as opposed to chats/IM) are best performed on desktop. On mobile they will always remain a compromise in usability and content quality.