After reading this it struck me that no revenue model could really be complete without some statement of the level of uncertainty. In science, for example, you would never make a prediction without also stating the uncertainty in that prediction. But none of the revenue model frameworks I've seen have included uncertainty.