Dunno that things have gotten much better over the years. End to end encryption is still fairly unusable to most. I think that rather than spending so much time on the cryptography we should instead work on a E2EE usability standard. Having everyone doing their own thing with respect to usability is not working out.
It provides the tools necessary to implement a simpler workflow.
It was pretty awkward to later emerge as the PGP expert at work. I was at that time a bit shy about explaining how I got so comfortable with it.
Otherwise, anyone who is using it at all will get court ordered and then, with our current system, you just lose.
Tor solves this by making traffic entirely homogenous such that any "it could be anyone" and by making nodes extraordinarily easy to host (such that any nerd can just leave one running 24/7). Even so there are way too few nodes.
But PGP-over-email is way too metadata leaky, and if you're doing anything that actually has a threat model, you need PGP-over-email to be itself common.
Thus, participants are pooled under 1 time constrained key exchange, participants sections origins are signed/time-stamped, and the ciphertext expires on a schedule.
Why? so at some future time side-channel key exfiltration can't be used to data mine the entire history of past tapped conversations/ciphertext. Even if a private signing key is stolen, it is difficult to rebuild the lost ephemeral key session.
Have a nice day. =3
Depends what you do, and if your organization has a persistent clown infestation. Some big business folks still use email as a human to API interface.
Running out-of-band public keychains is not something most people will ever need. =3