I mean it sincerely that, if you still don't see the 90-degree rotation despite producing a diagram that illustrates it yourself, then I do not know how to explain it better
OK i'll try this one more time, but i can concede my entire point is moot if light rays start downwards. I can concede that with what I now understand to be the correct EAT theory of the the suns light rays, then my problem doesn't occur. I'm completely happy with this outcome as it gives additional information to EA and provides a path for further discussion. But anyway, one final time for the folks back home:
Curves not shown, just "starting angle" and "ending angle"
There's no "90 degree rotation", there's a 45 degree angle of light, so the "visible portion" of the sun describes a perpendicular 45 degree angle...
I imagine what's happened here is you've taken your understanding of the sun "pointing downwards", and then seen my two points on a 45 degree sun, and my two points DO describe a 90 degree angle to your version of the sun... but in of themselves, they are not 90 degrees to MY understanding of the sun: they're just an additional 45 degree angle (45 + 45 = 90!)
So on my apparently totally incorrect version of the angles of light from the sun, then the top point would disappear first. This is now made almost entirely irrelevant, as you've explained that in EAT the light from the sun is almost completely downwards. Although we still have some points higher than others, they're also "further away" so you don't end up with the same problem of the area of visible light ending at the top before the bottom.
I think you'd still find if you modelled what the sun would look like as it disappears in your version, it would not be as perfectly uniform as observed