Pls provide us an example of how viewing a distant cityscape, with the closest subject 30 miles from us and our camera, providing ortho (as the goal is no distortions, right?), can produce 3d on our retina (either through a 3d viewer or with our unaided vision)?
He already talked about this point at least two times in this thread.
To paraphrase, he considers that this flat plane will be rendered at a distant apparent distance, creating a superior 3D experience.
I do not share this judgment (the second part), but it is the way I understood what was argued.
If you look at the projection matrix of the two cameras, the depth being on the denominator of the depth-dependant term, you obtain the same image for large distances (modulo a translation).
It is what we observe empirically and what make people say here that there is no interesting additional information with this kind of image.