A new problem arises is it looks like you can't see the moon over as much of the earth as you can see the sun. In the diagram you posted, the sun appears over a 12 time zone area while the moon seems to be visible over only a 3 time zone area (10:30 to 13:30). For example, it doesn't appear to me that a person standing over the 9:00 mark can see the moon at all (consider the same situation but with the sun not there and the moon being full).
The darkened part in the illustration is just the area of the moon's shadow upon the earth. Those aren't the moon's light rays. The darkened area just represent the
lack of the sun's rays.
The light rays of the moon are spread out nearly as far as the suns (a little bit less due to its lower altitude). It would be possible to see the moon from the 8am, 9am, and 10am locations. An observer at 8am is looking at the sun and moon and sun at different angles, and is not in the limited darkened area where the sun's light rays are obscured.
And an observer at 6am at the very farthest edges just sees the sun barely above the horizon. The moon has not yet risen.
Additionally, total solar eclipses are only 5 or so minutes long in any particular location. EDIT the above diagram seems to indicate Total solar eclipses last much longer than this.
The moon is passing in front of the sun at hundreds of miles an hour.
And lastly, can you tell if the cause of this "bendy light" is normal refraction as per RE or a new type of FE refraction?
I don't contend that the bendy light is caused by terrestrial refraction. The light is bent by the gravitation of the cosmos. The cosmos are pilling the light rays upwards with subatomic graviton particles too small to see or detect.
For the following question "wouldn't that mean we would all be sucked into the stars?", the answer is no. No one knows what a gravitation field even
is, let alone how light behaves in a gravitational field.
In the Round Earth model someone looks into a filtered telescope, sees the sun pass in front of a star - sees the star's light warped around it and proclaims - "ah ha!, since the sun is 93 million miles away and 900,000 miles in diameter, it must take a gravitational field of
this size to affect photons to the observed degree." But what happens when to the susceptibility of light in a gravitational field when the sun is really 3000 miles away and 32 miles in diameter?
I suspect that the real answer is that several million g's isn't necessary, and that light is just more susceptible to gravitational fields than previously thought.
No one knows how susceptible light is to a gravitational field. No one has been able to experiment with light and and apply varying levels of gravitational fields to it and observe its path. The only way we know how susceptible light is to a gravitational field is to observe the cosmos and make a grand series of assumptions. When we scale down the universe, we must also scale down those assumptions.