It's always funny to read optimistic statements that have since been proven a little too optimistic (I have a brilliant book from the fifties with the title 'Design for a Brain' - complete with electrical circuit diagrams) and The Cluetrain Manifesto which was of course a must read 2-3 years ago and not this year is an interesting case.

It's not that they are really wrong, it's just that the revolutionary tone really doesn't begin to describe how one feels about the present day internet with more and more intrusive spamming and less and less interesting new stuff.
But I shouldn't complain, and won't really. I prefer the 'brand new net' lovefest that Cluetrain tries to rekindle to the e-commerce babble one hears most of.

One thing is apparant from the enthusiatic descriptions of the one-to-one internet: The authors haven't fully taken in the 'power law' nature of the connected society described in Linked. One of the results of the power law structure of the internet is the notion of hubs - network hotspots that many network nodes find it useful to connect to. If you've ever been furtunate enough to functions as a local hotspot of some kind, you'd know that the gratifying experience of being useful is quickly replaced with a paralyzing sensation of not really being able to do much of anything except handle all of the requests streaming against you.
The dream of the one-to-one network where you can reach out and touch someone, and that someone can be someone who realy matters all the time is stifled because the hubs in interperson networks are very easy to saturate, i.e. paralyze intellectually by piling information requests on top of information requests. That's why there is such a thing as proper channels in a well-run organization. It is a device to liberate the connection points that all communication would run through to actually do something useful instead of just passing messages around.

There's an interesting equilibrium mechanism to this though, which could deserve to be properly mathematically modeled. Say you have a network, where the value of a node is the amount of information it is able to produce disseminate throughout an organization, and introduce prices for information production as well as for transmission.
Let every node try to optimize. What kind of network does those rules breed. Suppose the nodes are allowed to price their information production as well as their information dissemination as part of their own optimisation. What is optimal behaviour and what kind of network does the optimal linking behaviour afford.

The final conclusion is that the manifesto itself is a lot more interesting than the book about it.