There are several places where I feel proper regulatory oversight would improve the situation, however they can all be reduced to a single phrase.
Proprietary platforms that don't inter-operate are problematic.
Email is OK because like phone service providers anyone is free to run their own, and choose to use the services of a competitor while still inter-operating on a shared communications structure; much like actual postal delivery.
IRC is OK as a protocol (from the fair and inter-operable test), but was never designed to scale up to operate at this level.
Proprietary platforms, like Facebook, Discord, etc flavor of the week which have strong market dominance but which are closed are actively problems. That's an area where monopoly is clearly being used to keep out competitors and restrict innovation.
In Google's defense in this area they did try XMPP, and (as I recall) that experiment ended up as a failure since other platforms intentionally implemented a very poor baseline minimal interoperablity in ways that made communicating between platforms a very 'red headed step child' experience. That's why I believe whatever does come next has got to check all the boxes AND make them requirements of inter-operating, not 'extensions'.
The problem with XMPP is that it set Google's IM products back by a decade in terms of features. Google IMs still blow chunks.
It may now be possible to make an okay XMPP experience, but proprietary/custom protocols continue to be simpler to implement, better-performing, and less hassle overall.
We need something completely unlike XMPP if we want good federated IM to become a thing; something that isn't a bloated, exotic, "extensible", "well-formed", streaming XML protocol.
While I agree with your characterization of XMPP, it can still function as a gateway into richer platforms. There's no reason to turn off the lights. Chat hasn't changed that much in 40 years.
That said, if Google wants to earn some much-needed goodwill, it could work on a new federated chat protocol. Something binary and concise that handles all of the modern emoji, stickers, outbound call initiation, bots, polls, etc. And if you could use it to build all of {Slack, Messenger, Telegram, IRC, Hangouts} on top of it, that would be amazing. Especially if it were federated and had a reference implementation.
Commoditize your compliments, right? Anyone at Google looking for a project for promotion?
Don't forget iMessage in there. The social pressure to not appear as a green message bubble is intense and helps protect Apple's market share. They promised to open it up a long time ago, but of course that'll never happen voluntarily.
> Proprietary platforms that don't inter-operate are problematic.
The fediverse is coming along very nicely. Gab's joining of the ActivityPub network is a strong demonstration that you don't need to be part of a bubble to participate, and you don't need permission from anyone to join.
I think if Facebook, Google, and Twitter keep up what they've been doing, the fediverse has a pretty good chance of growing into something that can be used to reach sizable audiences.
I've been looking at ways to get my local, provincial, and federal government agencies to join the fediverse as well, as I think that's probably a better way to do public status feeds than relying on some opinionated Californian megacorp. I'm thinking maybe a slick publishing system that also proxies Twitter.
(INB4 “hurr durr spoken like a smelly nazi, flagged!”)
Much like postal delivery where I get a ton of u wanted flyers (spam) and the post office seems to help those make sure they get delivered to me rather than let me opt out entirely.
Proprietary platforms that don't inter-operate are problematic.
Email is OK because like phone service providers anyone is free to run their own, and choose to use the services of a competitor while still inter-operating on a shared communications structure; much like actual postal delivery.
IRC is OK as a protocol (from the fair and inter-operable test), but was never designed to scale up to operate at this level.
Proprietary platforms, like Facebook, Discord, etc flavor of the week which have strong market dominance but which are closed are actively problems. That's an area where monopoly is clearly being used to keep out competitors and restrict innovation.
In Google's defense in this area they did try XMPP, and (as I recall) that experiment ended up as a failure since other platforms intentionally implemented a very poor baseline minimal interoperablity in ways that made communicating between platforms a very 'red headed step child' experience. That's why I believe whatever does come next has got to check all the boxes AND make them requirements of inter-operating, not 'extensions'.