Don Norman wrote a piece arguing against many user-centered design approaches, or at least how they are thought of/about. I just wrote five pages in my notebook prompted by this, on the somewhat related theme of how what I seem to be pretty good at -- seeing and coming up with ways to help people deal with 'breaches in the canonicity of life' (Bruner) -- seems often to be hardly of much value in the world I mostly work in, where knowledge of and a feel for how things actually work (the engineering and operations aspects) is the main value. In other words, living within the canonicity and knowing how it works -- neither of which I'm much or uniquely good at -- is needed and valued.
How is this related to Norman's piece? Well, to me it says that having a feel for what will work (as Jobs clearly does at Apple) is much more important and valuable than following principled approaches that are meant to discover or critique what's wrong or missing from the dominant worldview, such as 'human-centered design'. There is a debate about this that I've been following in the interaction design world (this thread from the IxDA list is an example, where many are railing against the academic/research preference for 'user-centered design' as an approach, saying that intelligent/creative designers following their own instincts for what will work is actually more effective than all the techniques that have been developed for 'understanding users', regardless of how much more enlightened the latter is supposed to be).