Technology Review just released The Internet is Broken (note that this article is in three parts ... follow links within the article for subsequent sections). The article argues for a fundamental redesign of the Internet to support its uses today in an efficient and secure manner. The National Science Foundation is interested in addressing this question in its FIND project. There are two kinds of questions here, technical and economic/political.
The technical questions include: Why is a redesign needed? Wasn't IPv6 supposed to solve the problems addressed in this article? Why can't we continue to evolve the current approach?
To me, the more interesting questions are not technical. They include: How would you get existing users to switch to a new design (note the rather unimpressive adoption IPv6, some ten years after the standard was approved). Given the earlier items on network neutrality, do you believe that a new architecture would address Isenberg's wish list? Do you believe that the current infrastructure providers (AT&T/SBC, Verizon, et.al.) would go along with such an approach? How would a new architecture fare under some of the legislative initiatives underway in Congress?
Internet, Network Neutrality