On the Real Options yahoo list, focusing on the infoQ article written by Chris Matts and Olav we've had a great discussion about this very question.  As I understand it, a fundamental concept is to leave decision making until the last responsible moment. Also the idea of your saying that options should have value, and options expire are fundamental to this concept.  Chris and Olav mention starting with Black Scholes options forumlas from the financial markets as a method to help in decision making for software development. 

Currently in agile methodology, many groups use stakeholders to set the priorities for their features.  In my experience I sometimes suspect that the highest business value feature is not always selected using this method.  Especially in an enterprise situation, once the project has been approved.  In my experience, there is not a rigorous method to assist in determing business value once the funds have been approved.   

Corey Ladas of Modus Operandi, has a great piece on using Options in a Kanban that helps with selection method.  That has helped clarify some plaaning concepts for me in the kanban world.  But I still want a better way to determine business value.  In the email group we are discussing using costs, and wisdom of crowds for better decision making to get features from planning state, into the kanban.  So the question is, can we put actual value into our features using costs? Can we use mechanisms from social web to help us collect information from different groups?  I don't have the answers yet, but I feel those 2 questions lead us to the right answers.  If you are doing something like this to decide, I'd like to hear about it.