I think they're going about this the wrong way.

The primary problem they should be trying to solve is that it's fairly tough for a new user to join the ecosystem, get his friends to follow him, etc.

It's difficult for a new content-creator to get started. Much harder than with Facebook, LinkedIn, and even Google+.

Hashtags and @s are the features that make twitter such a valuable marketing tool. How do they plan to group the news and twitter trends?

If Twitter kills @, #, rt, mt ... that will be huge.

Good. Lumping together data and metadata has always been a bad idea.

I still say Twitter is a fad. Nothing it offers is a unique value proposition. If the right clique of celebrities moved on to some new, more fashionable network, that would be the end of Twitter.

Twitter itself may be a fad, but there's still a need for some kind of light-weight, small, broadcast mechanism.

If this had been done decades ago, we'd have an RFC and it would be a distributed protocol like email. Instead we have a single company dictating terms, for better or worse.

Maybe it's not too late? We need more protocols and fewer "platforms". Is anyone working on writing such a protocol?

The problem is getting somebody to implement and support it. XMPP has suffered because companies refuse to federate, and keep trying to build in their own proprietary extensions.