Systems engineers (for example the K&R "C Programming Language" book) didn't need elaborate math notation, like Knuth needed.
Eqn was invented precisely to do elaborate math notation. Integrals, sweeping braces, majuscule and minescule letters, suffixes, Greek..
Interesting to hear from the pre-Tex world. I'd like to see some examples of expert use of tbl and eqn; outside of manpages the roff ecosystem seems to have vanished.
The BSDs ship with a lot of the original non-man-page Unix documentation: the system manager’s manual, the programmer’s supplementary documentation, and the user’s supplementary documentation. These are written to be typeset with troff, not printed on a teletype like the man pages can be, so they tend to be a bit more refined. https://svnweb.freebsd.org/base/head/share/doc/
They are missing some chapters (eg eqn and tbl) due to copyright disputes but you can find them in TUHS archives https://minnie.tuhs.org/cgi-bin/utree.pl?file=V7/usr/doc
This http://www.nesssoftware.com/home/mwc/manual.php was set with the Coherent version of nroff. Result is damned excellent.
roff is probably very uncommon in these days. But not unheard of. A recent example is Donovan and Kernighan’s book The Go Programming Language which was typeset using an XML and groff toolchain. It was featured on HN a few years ago, see https://news.ycombinator.com/item?id=11470905
I think I "get" it better now. It's just at that time there was another school of more pragmatic engineering aligned thought in Unix using Dec10 RUNOFF derived concepts which morphd into t/roff and stuck there.
Maybe it's like Emacs and vi? Dec 10 SOS editor was like ed which led to ex and vi. If you walked into the TECO door you went lisp and Emacs.
To launch into Tex and latex you needed a professor leading you there rather than an engineer doing nroff and t/roff