Twitter Needs to Evolve

Imagine this: you go to send an email to a new contact you met at a conference and are told by your email server that you have to many email contacts and can't send this message until you delete some contacts. Or imagine that your all of your company's email is dependent on one external service provider, the universal email service provider for the world.
That's unfortunately where we are with Twitter and why Twitter needs to evolve, again. I say again because if you look at the history of twitter, it's already evolved and evolving, but it will need at least one more big step on the evolution chain. That Twitter history would be:

1. Twitter was launched primarily an SMS service for mobile phones. Although there was a web interface, the primary expectation was you'd get updates from a few dozen friends as text messages on your phone. I remember using Twitter "back then" around the Microsoft Mix conference in 2007 to stalk, er follow, authors and prospects via their Twitters to instantly know who was in the speaker lounge or playing video games in the hall. (This was in my old @jimminatel days before the @wrox account passed to me.)
2. Twitter's first big evolutionary step was to de-emphasize the sms/txt dominance and grow toward the web being an equal interface.
3. The introduction and growth of the Twitter API and various clients and applications based on it marks the current state of Twitter evolution. While many casual Twitter users still use the web interface, the leading edge are definitely migrating to client (phone and desktop) apps.

The Future of Twitter Is...
Twitter's next stage of evolution needs to look at the API usage as the future and build a service around that. That evolution probably shouldn't mean growing city-size data centers to house what will be en ever growing need for server space. Instead, Twitter needs to de-centralize. I see one of two possible paths here:

1. The first option for Twitter is to become something like e-mail, where there are common standards for internet exchange of email. Any entity that wants Twitter access would set up their own Twitter servers according to their needs or purchase access from others. Twitter's own corporate business models in this could be selling the first software to run these, or hosting a free service supported by ads, or selling premium access as has been speculated lately. This option would nicely allow for sites/servers running Twitter to take on domain names like the rest of the internet and for corporations to have corporate twitter addresses.
2. Another option might be more like many of the current peer-to-peer file sharing networks. It's hard for me to see how Twitter as a company can directly benefit from hosting and routing all Twitter traffic on servers they pay for. I can see a community of distributed p2p-Twitter services helping distribute this load.

Either of these schemes involves some fundamental rework of how Twitter works, and neither is a Trivial task. Neither completely solves the issue of how Twitter successfully makes money from their creation either.

But if Twitter becomes as popular as email, instant messaging, or text messaging (and why shouldn't it be that popular?) the current architecture seems unsustainable.