Just read this article by Ian Poole in IET Communications, which is a good overview of IMS architecture. (I don't know why I pay them subscription fees for the magazine when you can download all the articles for free!).
I am not convinced that the need for IMS stems from the desire to "converge". The main objective for IMS is to give the operators a mechanism by which they can charge for any user activity and control the content distributed over their network. With the introduction of 3G data access the operators realised that the future is bleak if they simply offered mobile data speeds for users to access the (mostly) free Internet content. Therefore they wanted a mechanism by which they offer an alternative or additional content they can generate revenue from and control what content the user has access to. Flexible content distribution platform and flexible service provisioning are by-products and not the core issues.
With IMS revenue generation may not be focused on the end user. It could well be based on a sharing revenue with 3rd party content providers or by selling to marketers or retailers alike access to customers.