... is it of any real use? Wouldn't any real task, perhaps except evaluating architecture, be supported better by a more abstract and compressed view? To me it looks like this thing suffers from the same problem as almost all 3D virtual worlds: It pretends to be real by visual appearance, but isn't in any way. It isn't as natural as it tries to be, because you cannot act as if you were there.

Now the aim of user interface design is to let a user perform some task in a more efficient manner than it would be possible without the software. For which tasks could this be true in this case? I don't see even a single one. And no, putting virtual worlds on the virtual screens wouldn't help much.

It seems strange, but NYSE is very physically based, it's not nearly as automated as you might think--everything is based on people scurrying around from computer screen to computer screen.

Ok, but it is still questionable whether such a literal view is really needed, or even supports the task. When I'm trying to locate a street, I won't use a literal 3D representation of the city. I will use a map instead.

German TV station ARD some time ago introduced a plaything to their weather forecast. It's a flight over Germany from one randomly selected City to another randomly selected city, showing predicted rain, blue sky, etc. I consider this the most useless part of the whole forecast. It does not tell me anything about the weather to be, and in particular nothing the 2D maps before and afterwards couldn't show.

The second post was pretty accurate, that you have to think of this as a physical system with a weird 3D structure. It's like designing an airplane or car, blueprints are nice but it's also nice to see a model of the whole thing.

In general, though, I completely agree with your attitude--3D is almost never useful, it's good mainly for representing things that exist in real, 3D space and have a complex 3D structure.