As much as people are writing and thinking about the net, what you really need to read to understand anything about the net is
- Some information theory
- Basic cybernetics - which is hard to come by without having to read a lot of gunk by the way
- A little McLuhan
- A little microeconomics
- A little anthropology
And that's basically it, as far as I currently see it. Then what you want to do is extract out of these fields a few salient stand alone models to reason with, and then run these models over a good set of pivotal examples of net phenomena.
As the internet continues to dissolve it's becoming less and less appealing to study specific internet literature, and more and more appealing to study the fundamentals of the fields that describe the powers that drive the net.
For extra credit, teach yourself some of the basics of the theory of computation, to understand exactly how malleable models are. It's way too easy to set out to discuss the internet, and not realize the central current powering it all - which is the extraordinary capability given to us by digital machines - of making any abstraction a reality, and in turn move our thinking to this new platform. Thermodynamics tells us that perpetual motion machines are impossible - but that's only in reality. In the realm of ideas we truly have perpetual motion machines, and you're reading these words using one, right now.