I'm pretty sure that is an artifact of the fact that this is a blog post--it's easy to think of a hypothetical example of an innovative product in the MIT mold

I'm pretty sure that is an artifact of the fact that this is a blog post--it's easy to think of a hypothetical example of an innovative product in the MIT mold ("existing production process, but much better/cheaper!") and difficult to give a hypothetical example of an innovative product in the Stanford mold. We know we want to produce things more cheaply--we didn't know, before Facebook, that we wanted Facebook. As pg puts it, writers who need to come up with startup ideas "come up with an idea that sounded plausible, but [is] actually bad."