Pictures of RE have to be dismissed as part of the conspiracy again!

I was doing a bit of thinking about the FE appearing round from high altitudes due to EA, when I realized something. While it would appear spherical, the sphericality wouldn't appear as it does. Let me explain this a bit more in-depth.

Say you're floating at an arbitrary point above the Earth. Someone at an arbitrary point on Earth shines a flashlight towards you but remains parallel to the Earth; that is, they shine a light along the intersection of the plane of the Earth and the plane containing them and you perpendicular to the Earth. This *should* remain parallel to the Earth, given the exponential nature of EA. This is going to intersect the circle centred at them and with radius = the distance between you at one point.

Now have them shine a light straight up. Similarly, EA won't affect this trajectory, and so it'll continue along a straight path, intersecting the same circle at a different point. Notice how you're between the two points of intersection (well, I suppose you'll be between them anyways, but on the minor arc.)

By the Intermediate Value Theorem, there should be an angle between these two they can shine a light at such that the ray of light hits you.

This means that, if you happen to look in the right direction, any point on Earth should be visible from space, assuming you're high enough that obstructions don't get in the way. Thus any land mass on Earth should be visible from space.

However, Asia clearly isn't in the following image:

(EDIT: I didn't realize it was obnoxiously huge, so I'll link to it: http://eoimages.gsfc.nasa.gov/images/imagerecords/57000/57723/globe_west_2048.jpg)

Thus this image, and most others, are again part of the conspiracy (assuming a FE) and not the result of EA.

Did I get anything wrong?

yes this image is obvious cgi/paint ... u need to use a real image

Yes, I believe you have gotten something wrong. Let's look at the assumptions being made, shall we?

1. EA will not bend a light ray parallel to the Earth

2. EA will not bend a light ray orthogonal to the Earth.

The first assumption is incorrect.

Now, it's rather difficult, imo, to visualize EA. Let's use refraction as a metaphor of sorts, since they work via the same process.

The equation to refraction is nsinθ=nsinθ. Let's set arbitrary values for the ns. 1.33sinθ=sinθ. This happens to be a light ray moving out of water and into air, but that specific doesn't matter. I just happened to know the n values for those two. Do you see that the light ray will still bend, even if it is parallel to the surface? 1.33sin90 /= sin90

This analogy with refraction is highly misleading, since in order for refraction to happen, light should change its medium. Now, if it's travelling parallel to the line between the media (plural of medium?), it will obviously not refract, since it will remain on the same medium. The angle theta is measured to the normal of the surface, so if theta=90 it means that the light ray is parallel to the surface and it will stay on the same medium. If it comes perpendicular to the surface, theta1=0, so obviously theta2=0 also. (sin 0=0)
This is different from EA because EA (as far as I know) is a constant force pushing light upwards and it acts always, independently of the medium light travels. Anyway, I think a light ray parallel to earth would still be bent upwards by UA. So you were right, but the analogy with refraction was faulty.