I understand not knowing what lwn is… but why insist?
"The manual pages in this project document primarily the Linux kernel and the GNU C library, but also consider relevant differences in other kernels or C libraries. These pages are most of the section 2, 3, 4, 5, 7 man pages for GNU/Linux. A few pages are provided in sections 1 and 8 for commands that are not documented in other packages, and there are a few pages in sections 5 and 8 for the timezone utilities."
Most manpages for individual command line programs are written and shipped by the projects or authors of those programs. And if there are manpages for functions in non-libc libraries those are generally provided and shipped by the authors of those libraries.
The wide range of origins for manpages means there isn't going to be a single consistent style over every manpage in a Linux system (and of course "don't ship a manpage at all" remains a popular choice).
As far as GNU tools go, their attempt to move docs to (tex)info format for better interactivity has largely failed, so I guess most of that content is slowly moving over to manpages now.
I think the only community where it kind of succeeded is the Emacs community, thanks to the built-in info reader and that a lot of Emacs docs got written in info.
The one for traceroute(8) has some great commentaries on the sample output ("God only knows what's going on with 12."). It's a personal favorite.
I dread Linux man pages
LWN is not a documentation project. It's a news outlet.
What I mean by "unbiased" is neutral coverage on actual news, meaning just the facts. If you pay attention to the coverage at LWN, there is an inherent Rust bias that dominates regular news. Just try to go more than a few days without new articles being published with a pro-Rust stance. Actual data with stories and coverage over the years would illuminate this easily.
>unbiased