This is a question that rises to the level of heresy in the industry, akin to asking if everybody should be allowed to have a PC.
Nevertheless, the question was asked in this recent article in Application Development Trends. "Building the ROI case for EAI is a challenge. The lack of integration hurts, but it is hard to quantify. The problem? The benefits of EAI do not readily translate into quantifiable business gains. How much is it worth on the balance sheet to get new functionality working a few weeks sooner? Whats the gain in hard dollars from an improvement in customer service?"
These are questions that need to be considered before companies can be convinced to pour money into Web services and service-oriented architectures. Is faster response time from multiple systems a nice-to-have, or a need-to-have? I know of a company that just converted from DOS to Windows. Yet, their business brings in gobs of money of every year. I don't get the impression that only using DOS workstations all these years held them back in any way.
I'm afraid the answers aren't so clear cut as they were when we were replacing index cards with computers. Web services and SOA are not about technology; they're about what they will do for the business.