Personally, I just think that Greg doesn't want to admit that he is wrong. Logical arguments have been presented that he has rejected, and he wants to move to perception that can't be measured.
To help him out, let me propose a theory. I don't know if it is true.
Fact: Our eyes are designed to take a trapezoid image and convert it to a rectangle.
Fact: A flat screen will best represent that image over a curved screen
Theory: When presented with a curved image of a straight line, might our eyes interpret the image in such a way that the image appears to have more "depth"?
Basically, when our eye sees a straight line when it is expecting a curve, it adjusts the image into our head to think that we are seeing something that is not is deeper than a flat image? Some people might see that additional "depth" and like the picture quality more than a flat image. It isn't accurate, but it could be like the concept of 3D - you see a fish right in front of you even though it isn't really there.