Chapter 2 of the book is about the next-generation network and IMS. I extensively revised and added to the chapter today and was struck by how we have to move the IMS debate on.
It is NOT an argument against IMS that IMS is complex. Lot's of things are complex (aircraft, GSM, legal systems) but they work and are manageable.
It is NOT an argument against IMS that carriers may use it to destroy their Internet competition. They don't need IMS, a Narus box will do perfectly well. Regulation, and to an extent market pressure, are the forces which need to be applied to anti-competitive carrier behaviour.
IMS integrates many of the functions carriers care about: authentication, authorisation, generic access to service platforms, security, access efficiency via compression, roaming, sophisticated call-control, QoS management and most importantly metering and billing.
If you want to attack IMS, whinging about its general unpleasantness is not the answer (honestly, it's quite interesting when you get into it!). Look for paradigm-breaking services which need signalling/session management functionality not found in the IMS roadmap. If the service is popular, then you will get market share and the cumbersome IMS standards process will take 3-5 years to change direction and absorb your new functionality.
No, I don't have a view as to what this new functionality is either, which kind of weakens the argument against IMS, don't you think?