Alright, i'll try to reply to both your posts, but i'll have to use my berserker rage, as Chris Claremont might narrate it

>modern internet puritanism
>I suspect this may be caused by either the migration of insane tumblr users to the wider internet after the sitewide ban of pornographic content, or the influx of "new?" users
The first thing i believe is that it's not really moral prudeness. At least not as we conventionally recognize it.
I believe this trend started on Tumblr, as counter-culture to the edgyness that followed the 2000s silly irreverence. 4Chan, for example, used their in-jokes to poke fun or subvert a topic, but still based on traditional views of what's tabboo, while Tumblr harvested the people that were directly opposed. To illustrate: In 4Chan you could have "traps aren't gay lol (why would anybody think that)" while in Tumblr the average user might directly be against homophobia.
The Tumblr mindset grew partly out of the whole anti-SJW movement being butthurt about them all of the time, and the flourished with the Twitter migration

All of that said, i believe this neo-puratism starts from a place of legit concern and avoiding to be offensive (trigger warnings and such), gets propped up by reactance to the edgelords and then grows deep roots due to cancel culture and YT drama. Towards the end of the 2010s drama and cancel content (God, i hate that word) put forward the idea of internet judging firmly on the table. Suddenly internet personalities clashed with each other not necessarily on moral principles, but just generally on dislikeable behavior, and it was clear from the turn on public opinion, clout (Hate that one too) loss and, in the most notable cases, public apologies that came after that this was a tangible instrument of the public opinion. It is no coincidence either, i think, that during these age of "reaction", "drama" and "commentary" youtubers, the word cringe started to take off. There was an ambient of peer judgement. In hindsight IDubbz's Content Cop series might ironically have been the most contributing to this, since at the time it was seen as one of the cool guys doing it.

Notice how things like "not a good look" or "It's unhealthy" is not a direct damnation to the poster they're replying to, but rather, rely on letting the digital bystanders know that they don't approve of what their seeing. To this you could add how people now (Of my generation. I'm not going to pretend to be an 0ldfag) tend to have less barriers between their online activity and real life, but i digress


>Replacing words with "safer" alternatives
This is was i was researching in the first place. The short version is that it started as a way to not trigger people on Tumblr. In McCulloch's book there's an Tumblr post quoted explaining why it wasn't of any use in the first place, since all it did was bypass wordfilters and it's not like it's hard to decipher what f*ck means.
Nowadays, unalive, seggs, PDF File and such are ways of working around censorship algorithms. I have a whole script talking about all of the above and ranting over people bending over, letting their vocabulary be dictated by a program meant to appease advertisers, however, this post has gone long enough