So what I'd ask of you is whether you can recast the book in a way where you see it as defining an underlying problem which we might agree on and then as search for remedies. This is my take. You might have a different one. But perhaps the method in my take can be incorporated in your own approach.
The underlying meta problem that cuts across most of the examples we've considered is "myopia" in decision making in real word decisions. This means the present payoff is overly stressed in determining the course of action. In contrast, a "farsighted" approach would entail a full incorporation of future payoffs as well as the present.
If that is the problem, then the next question is whether it can be addressed by modeling decision making as fully rational. The issue here is that one gets either into tautology quickly or an argument that really there is no problem. The standard rational explanation for myopia is that personal discount factors are low or personal rates of time preference are high (the discount factor is 1 over 1 plus the discount rate) so the future has little value for people even when it is fully accounted for.
The behavioral approach moves away from the fully rational approach. In Nudge we've seen multiple possible explanations - procrastination, inability to deal with complexity, and following the herd to mention a few. The benefit from a modelling point of view in bringing those explanations in, recalling our discussion about Mill where keeping the model simple is a virtue, a point I subscribe to as an economist, is to enable that myopic behavior need not be optimal a priori. And if that is true, then it becomes more of an empirical question whether myopic behavior is optimal or not. (You may recall in our class discussion some of you argued that procrastination in your schoolwork was optimal because you could be focused and intense when doing work near the deadline. I believe that is a only a rationalization and you'd be much better off in the learning by working well ahead of deadlines to produce your deliverables.)
In cases where it can generally be agreed that we are getting myopic behavior that is not optimal then the next question is what to do about it. The discussion at this point becomes something like the economics of market failure. In that literature, typically solutions are second best. And there is some disagreement within the profession on whether such solutions should be implemented or not. One can make a case for laissez-faire even in the presence of market failure because a regulation cure is worse than the disease it is trying to remedy. But one might also muse about why markets don't seem to come of with a cure themselves. If there is a problem shouldn't there be a profit opportunity in fixing it?
Sunstein and Thaler obviously aren't pro laissez-faire. Some of you seem to be. To me that shouldn't block learning from what they say even if doing so is a bit of a struggle.