Most of his failures on that list seem to be a disconnection from social trends and/or a misunderstanding of the ways that people will actually want to use technology, as well as overestimating adoption rates. He also seems to think that as soon as all the component parts of a device have been made commonly available, inventions that combine them in straightforward ways (i.e. GPS navigators for the blind) will appear instantly. In short, were it not for sociology, he'd be doing a lot better.

Then again, my understanding of the way he does business is to fill in those gapes himselves. The products his company makes seem to be focused on combining new technologies, and he encourages people to do the same.

He was wrong about voice recognition, but once that's there, suddenly about a third of that list will fall into the same category: just a matter of overestimating the initiative that the market would take. It looks good to me.