What struck me most about this paper were its strong assertions without any data to back them up. I was left wondering if what the author stated really worked, or if it was just what he found to be true.

Overall, I am inclined to agree with his premises. Seeing "real world" models of complex algorithms helps me to understand them. Andy on stage ripping a phonebook in half to model binary searching, Spike playing "the remainder game" on the board to emphasize induction, and many other "trivial" examples really help comprehension wise. It is much easier for me to tackle a problem when I can relate it to something I understand. I thought this paper had great potential, and hopefully someone can provide data to back up its assertions.