the fact that to get to the other person, the sun doesn't go AWAY, it goes DOWN BELOW the horizon absolutely discredits your point
have you studied basic perspective modelling? it's the same principle that cuts the botoom half first of a ship/whatever at the horizon : DIVERGING lines of perspective after CONVERGING ones
Yes, I've "studied basic perspective modelling", especially the
Rowbothamesque version of "How to lie with perspective.".
Quite ingenious really, isn't it? But, the
Rowbothamesque version seems to be all you've looked at.
I can't make any sense of your "DIVERGING lines of perspective after CONVERGING ones".
If by that you mean perspective lines diverging
after the vanishing point, rubbish!
The vanishing point is where? Not necessarily on the visual horizon - there are plenty of photographs verifying that!
In actual fact each object has its own vanishing point, depending on its size (and contrast actually), not in any way related to the visual horizon - that assumption is just a drawing aid.
As a drawing aid, putting the vanishing point on the visual horizon is convenient because changes in distance get greatly compressed in the distance due of course to perspective.
For example, 100 m of distance 1 km away takes up 10 x as much image space as 100 m of distance 10 km away, and so on.
Your diagram, however, isn't a bad illustration of something disappearing over a curve, even though said curve is approximated by just two straight lines.
Then the photographs are are a good illustration of a person being hidden, feet and legs first, as he walks away over a piece of slightly curved ground.
In other words, without corroborating evidence, I do not believe that those photos are taken over flat ground.
And they used to say that the camera never lies! That's probably true, but the operator can certainly take intentionally deceptive photos.