He argues that Enron by and large made good on its fiduciary obligations to disclose what it was doing (it provided all the pieces of the "puzzle") but that the scandal was that nobody could figure it out (it was still a "mystery".) The "mystery"-"puzzle" distinction is not one I buy wholeheartedly.
But what struck a chord with me is the idea that the deals Enron was striking were, while precisely defined, unclear in effect. This is very much the sort of situation in which you should pull out a computer model and play around with it hoping to get some insight.
My advisor would occasionally talk about the virtues of "compilation" and speculate about whether it could be applied to generate/model-check scientific papers, or even law or legal decisions. I'm somewhat dubious. but it seems to me that this sort of complicated financial agreement would be a nearly perfect match. With a formal definition of what the contractual agreement is, a model could be compiled that would allow any interested parties to try to figure out what the results would be under various scenarios.