The main problem I have with this potential business model, is the overt focus on adding friends as a part of the revenue stream. I think the focus should instead be on 'what we can do right now, with our userbase to make money?'. There is a large pool of current users that would gladly pay if twitter had attractive premium features. Maybe they should focus on extracting revenue out of their existing user base first and then move on to a more dynamic business model second?
Yeah, and then what happens when Burger King comes out with something that challenges your revenue stream that is based on making connections by encouraging you to break connections?
Alternatively, if the monetary value of the network is defined by the edges, rather than the nodes, and everyone is connected to everyone else, how do you determine the valuable connections from the meaningless ones? As the number of connections goes up, their individual value would seem to go down. Beyond that being basic economics, you're setting yourself up to be limited in the kinds of additional things you can do as the meaning behind the individual connections approaches zero.
If twitter added some metadata behind the ability to create connections, like allowing users to group their connections into buckets (initially preconfigured to encourage meaningful use) like "co-workers", "friends", etc and let you filter what you publish and what you subscribe to in different ways, it could work. But then they'd have created FriendFeed.
