Gopher was more widely used that WWW until 1994, and it was a basic install on most Internet nodes in the first few years of 90's. The point of Gopher was it was one of the first widely deployed protocols that allowed you to find things, like a Mosiac Browser to download. It died quickly but was an important stepping stone.
In 1990 IRC had average of 12 users :) (but other chat programs were used before IRC) . True FTP, other chats, MUDs, email, etc. were widely used/available.
Windows 95 had a native TCP/IP stack. Lan Manager had it from 1990 (a bunch of that expertise came from hires from Finland). And Windows NT had it in 1994. In '95 everyone was using TCP/IP and pretty much was standard at M$. I was there. It's true the browser didn't kick into high gear until '97 after Microsoft decide set-top boxes might not be information access solution of the near future. Microsoft was a little slow to admit that the Internet might win but eventually they came 'round and put some effort into it. Interestingly, both Gopher and Mosaic came out of Midwest universities in early '90s - great stuff, fun times.
Interesting idea, but I'm not sure I agree. I was doing a lot of telephony/mobile back in those early 1990s and there was a huge amount of formalism. We tried to model everything as FSM that transitioned states based on received messages. We even used code generators to generate code directly from diagrams. I believe Erlang grew out of that same approach. And initially, systems that failed would simply restart by default.
Then the idea of adding a supervising and observability was a natural addition to these systems.
That said, I think OTP is a great tool for modeling failure domains and I think it does a great job.