The goal of the experiment was not to measure irrationality. The goal was to measure neuronal activity involved in decision making, while fuzzing the decision making system with too many options.
Maybe I'm just broken, but this is a long way from how I choose a choc bar, from 3 or 20. My choice will change depending on what I had recently.
I don't have time to skim the actual paper, but I'm more with the GP that it seems a pretty poor experiment.