I spend a not insignificant amount of my work week now dealing with quarantined or bounced mail from other companies that can't or won't set up DKIM, SPF, and DMAC properly. Not individuals, companies, with IT departments.

You'll never get most people to properly manage their own email server.

This is true.

And it's a problem with email itself lacking a native security infrastructure.

I don't know what the answer is - maybe less reliance on email? But then it would have to be supplanted by something else.

Seems like we're still in the developing phase of all this stuff.

Email grew during a time when connectivity was sporadic. I'll check my email when I get connected. So store-and-forward.

We still need the store-and-forward capability, but we now rely on instant delivery.

Then there's the conversation vs letter idea, files/attachments, etc.

Corporate systems try to combine it all, which makes sense. No reason to move files about, instead have a repository and send coworkers links to the files since we're all part of the same infrastructure (it's a database after all).

If we look more abstractly at the major functions people use, there's largely messages (ad-hoc, one-off), conversations (something like messages, with group management, longer-term chains, etc), data sharing (files, images, video, preferably links so people retrieve as needed), meta-data (say contact info, business info, location data, etc), and who's know what else I'm missing. Designing system/s to manage all this while being extensible seems the big challenge.

To go full circle on this, even if everyone self-hosted this same repository/messaging platform, we'd need some kind of federation capability, with security and trust management.

It would be interesting to see any research on this from Microsoft (and other groups, universities, FB etc) - their R&D org really knows their stuff. I'd like to see the high-level, abstract, major categories of elements.