WSs aren't a magic wand, they just allow you to do more with the same.
You basically end up with a half-assed, buggy, and potentially insecure reimplementation of HTTP running over web sockets running over HTTP at that point.
Are you going to do session management and identification by issuing a unique identifier as well? Is that unique identifier going to be valid in perpetuity or have a an expiry date? Is it going to be valid and secure across all sub-domains or will it be prone to cross site scripting attacks?
The purpose is in the things explained well in TFA: single source of truth, no replicated state, simpler development, and so on...
Who said the purpose would/had to be in handing "the server caught fire" scenarios?
It works for certain things and not for others. So, it's not "the future of web software" like the title says, it's one additional possibility for when it makes sense.
Because you can do things stateless HTML alone cannot do.
No one is trying to re-implement HTTP, or even remove it from the equation.
What people are trying to reimplement (or rather, replace) is the HTML + JSON:API + React + Redux mess. Websockets is a means to that end.
Unfortunately, great intentions alone won’t make this work; someone has to actually answer the hard questions about how it’s going to work too.
All the answers I’ve seen so far trying to answer them have been vague hand waving...
That doesn’t leave me with much confidence.
People did better than that, they already built applications with this model.
Here is the documentation for Phoenix channel authentication (what Phoenix Live View is built upon): https://hexdocs.pm/phoenix/channels.html#using-token-authent...
Additionally, here is some documentation on the security considerations specific to Live View: https://hexdocs.pm/phoenix_live_view/security-model.html
Searching can probably find you more information.
A crash or someone hitting refresh on his browser is not that different in this respect.