In "The Paradox of the guided user: assistance can be counter- effective," van Nimwegen asked two groups to perform the same tasks. The first was allowed use a computer; the second group only got a pen and pencil. The second group executed all tasks faster and performed substantially better. In addition, their solutions to complicated problems were more creative.
Van Nimwegen says much software turns us into passive beings, subjected to the whims of computers, randomly clicking on icons and menu options. In the long run, this hinders our creativity and memory, he says.
I don't think there's anything wrong with allowing people to format a document without exerting too much cognitive load, and I don't think that using wizards to guide users through otherwise-complicated software tasks will turn humans into a bunch of Wall-E-style consumers. They may learn less about the software and have less situational awareness about the task (writing documentation?) they're completing, but doesn't that at least create the potential for users to spend more time doing value-added work instead of writing administrative documentation? I don't want to waste cognitive capital on writing a report; I want to spend that cognitive capital on thinking of new solutions or connecting disparate pieces of information, things I wouldn't use Word for anyway.
I think the lesson here is that interfaces should not be designed to take the user out of the equation entirely, but rather to keep the user engaged in order to optimize the human value the user brings to the equation. So keep the boring, administrative, and easily-automatible stuff out of the hands of users and use the interface to facilitate rich, value-added interaction.
Image from here.