Dean Leffingwell has a great post on Prioritizing Features in Scaling Software Agility. He argues that, since value delivering is a key economic driver, basing prioritization on Cost of Delay, thought not all that different from prior models, might be a very efficient way of working the backlog.
The article is very well researched and makes a compelling argument. So much so that I will make some proof of concept using the suggested method. We work wih stories and so far, we have "enhanced'' the business value of our stories by adding Hidden Business Values and R&D values. That way, we don't expect our PO to prioritized the stories with a heavy "back end" value (we call "hidden such as logging, parallelizing, etc) or corporate concerns (in our case, R&D because we have strong R&D programs in Quebec). By adding the Bv, HBv and R&Dv, we end up with a Total Business Value, which we use with the estimated Effort Value (Ev) for prioritizing stories.
I wanted to use a metric regarding risks. It seems to me that monitoring Cost of Delay metrics would also be a very efficient addition to the way we currently do things.