I think his idea of "programmer" is closer to "mathematician" --- and indeed if you think in that manner, functional programming looks perfectly natural and obvious. The problem is how many of the programmers - and by that, I mean anyone who writes instructions for a machine - actually do think like mathematicians; and in my experience, not as many as the ones who can easily grasp the sequential, imperative model of machine execution. Not to say that thinking of computation as a function that takes an input state and produces an output state is conceptually interesting, but I think that's a bit too abstract for a lot of people.