>>51
The issue here is that you're too dumb to see where he's coming from. When he actually tried out these ideas in a product, they worked great. He extrapolated according to the needs of a more modern system, and I'm sure his interface would work great for certain applications. I suspect that since the amount of local data we're likely to deal with has multiplied a couple million times, not to mention everything on the Internet, this wouldn't scale. But since this UI was never completed (the people carrying the torch don't seem to be a crack team of programmers), no one can say how well it would work out. Especially not you. For all we know, it would be every bit as good as he thought.
Anyway, here's how science works: a researcher with his team develops a theory for several decades until the experiment is finished. You might not call UI design science, but that's just why I don't have any respect for your opinions. In this case, the experiment was never finished, and the results are up in the air. Modelessness in general is pretty nifty.