One of the ways to think about the complexity of SOA is as a multi-sided platform.
Some of the pioneering work on multi-sided platforms has been published by Andrei Hagiu at the Harvard Business School. (Follow links from his homepage to some of his recent papers.)
In November 2006, I posted a brief commentary on Two-Sided Markets, referencing his work, and indicating its relevance to a service-oriented business strategy.
As yet, relatively few people have made the connection between multi-sided markets and SOA. In June 2007, Richard Friedman posted something useful on Multi-sided Platforms for Business or Software, and I found a paper entitled Optimizing the Supplier Selection and Service Portfolio of a SOA Service Integrator presented to a 2008 conference by a group of German researchers.
One of the attractions of SOA is that it should allow people and organizations to collaborate more effectively and cost-effectively. From the supply-side perspective, this means collaboration between the users of your platform. Your platform may or may not support a rich and open variety of collaborations, including mashups and other third-party initiatives, and this richness and openness significantly affects the demand-side value of your SOA platform.
The critical economic question here is not the economics of scale on the supply-side but the economics of scope and agility on the demand-side. Thus the critical question for platform design is how open/closed these platforms are, and the extent to which they constrain (overdetermine) or enable (underdetermine) demand-side activity.