There have always been central agencies that were quite important to the Internet. The text even makes a reference to it: Jon Postel is featured prominently. He, personally, and the Internet Assigned Numbers Authority shaped the network in a big way. The Internet Engineering Task Force has been instrumental as well. The text speaks of the quality of RFCs as the defining documents of what the Internet is today — but it requires a lot of dedication, editing and work to have a collection of standards in such a high quality. Then, there’s the local RIRs: ARIN, RIPE, AfriNIC. For all practical purposes, they’re centralized structures as well.
Don’t get me wrong, I’m for the power to innovate on the network. Innovation is a good thing. I’m also not against making money on the Internet; heck, I’m trying to make my living off networking stuff, too. But I’m against such innovations like SiteFinder by virtue of which Verisign broke the DNS because they want to make more money. I’m against reforms that would remove any form of netizen-based democracy on a global scale that can ban VeriSign from implementing that service again.
Also, it seems that all the text is concerned with is the domain name system. It’s true that it is an important part of network infrastructure — and a publicly visible one at that — but it shouldn’t be equated with all internet policy. There’s other interesting policy decisions around; how to switch from IPv4 to IPv6 for instance. But that has nothing to do with registries or registrars.
Maybe the author needs to do a little more research?