I liked the quote (from Craig Larman?) about the value proposition and would like to extend a little on that concept.
To my company, the value proposition is not so much the UML tool. We are a IBM Rational customer and have been for some years. While Rose really lacks in terms of functionality, what it does it does it well enough for our needs. Moreover, we integrate it with Rational RequisitePro and Rational SoDA for requirements management and documentation automation. Since we use a development process, what we end up with is a software architecture document that contains what we need to successfully develop a system--or subsystem or component. The value proposition from IBM Rational is the tool integration. Don't get me wrong: Rose is not as complete as EA, but is complete enough. I remember a methodologist saying that 80% of UML is class diagrams sequence diagrams and use cases, and that's what we use. Ocassionally we'll use a state diagram, or an activity diagram, but it's definitely not the rule.
After I learned about EA, I tried to sell its value proposition to the company, but price is not everything. It would be prohibitely expensive to maintain our current models, train personnel, etc., should we migrate to EA.
Now, going back to your original question--although it seems that you've already convince your boss--what is the value proposition from Enterprise Architect? To me that's the question to answer. It does not matter if something is $100, $1,000 or $10,000. For some companies, $10,000 for a tool in the large scheme of things are peanuts if you're going to be making 10 million. For other companies, paying $1,000 is too much.
So, my recommendation is to take a look at what the tool does and how well it would accommodate your needs.