did you read that survey? i was struck by the fact that a 50+ page survey of future internet architecture research did not have room to identify what problems we have failed to solve in the current Internet. on the first page Raj relegates this topic to an endnote reference, which links to this slide set: http://www.cse.wustl.edu/~jain/talks/ftp/in3_hu.pdf ,
which is a different (angle and) list than i found by interviewing engineers and operators back in 2003-4: http://www.caida.org/publications/presentations/2005/topproblemsnet/ . both lists are worth discussion, but do we have any document that expands on the problems we're trying to solve, what has been tried to solve them in the past, and why those solutions have failed? as lixia and dan reminded me, in any other field of engineering, such a document would be the first phase of R&D.
i deepskimmed protocol politics, i can recommend it as a take on ietf history, but that was only part of what drove internet protocols (it's even a subset of the politics that drove them). her advisor's book,