I wish i had this back when I was teaching gen-ed science courses in college. I tried to do something similar, but at a much smaller scale. Some random observations, that would help flesh the content out:
A big reason “type 1” reasoning is so often wrong is these decision making modules evolved under very different conditions than we currently live in.
I always liked Pinker’s description (from “How the Mind Works”) of the nature of the conscious mind by reverse-engineering it: it is a simulation of a serial process running on parallel wetware, for handling problems of astronomic combinatorial difficulty. This relates nicely to the cost of type 2 processes.
The modular nature of the brain and the conscious mind’s primary role as a confabulator in tying together type 1 responses is nicely highlighted in Ramachandran’s books, especially in his descriptions of hemispatial neglect.
You start the piece with some nice specific examples of cognitive errors, which gives the beginner reader an idea of precisely the sort of mistakes he/she can avoid by learning more… a great way to stimulate interest. Considering how important ‘type 1’ and ‘type 2’ processes are for the rest of the piece, I think it would be improved by beginning that section with similar concrete examples. Catchy names other than ‘type 1’ and ‘type 2’ would also help… maybe ‘Colbert’ vs. ‘Stewart’?
Make that ‘Colbert’ vs. ‘Spock’ :)