The plane's wing is curved, only its curved upwards.
When a plane is in flight its wings aren't perfectly straight. Airplane wings (even on a jumbo jet) are quite bendy. In flight an airplane's wing can bend all over the place.
Watch this: http://www.videosift.com/video/Boeing-777-Wing-Stress-Test
So sounds like we all agree the wing is bent upwards very slightly because it's made of aluminum and it's holding up a plane. That bending is real, not an illusion caused by any curve of the window. I pointed this out in my initial post of the picture.
Nice side stepping of the issue, but to continue with the logical part of my question, if the wing is actually curved ever so slightly, but the horizon is curved much more, in the opposite direction, even though both the horizon and wing are seen through the lower half of the window, then how is the window causing this illusion of a curved horizon?
It would appear the convex glass theory fails to provide adequate explanation of the observation. Besides I could just as easily argue that a convex glass was working in the opposite direction of the curve of the earth and partially flattening it. Without measurements of the glass and its optical qualities at that particular angle, how can you know what its effect would be?
It can't be that a convex glass will always cause something flat to appear curvier but never cause something curved to appear flatter.