In regards to the setting sun, the refractive effect of looming is not taken into account by "modern science"; the heliocentric model is assumed thus a refractive effect causing magnification due to an atmospheric plane is excluded from consideration. However when the scientific method is applied the refractive effect of looming can be experimentally observed to occur over relatively long distances and must apply to the sun.
First, looming is taken into account with the setting sun. It is what causes the sun to appear to flatten out as it sets.
I assume that you are using looming/refraction here to explain why the sun does not shrink as it gets farther away, but I think that your explanation raises more questions than it answers.
Can the effect be enough to account for the 3x magnification that would be necessary for the sun to be the same size when it is overhead and at dawn and dusk at the equator, and the 6x magnification necessary at the southern edge?
Why aren't other objects magnified like this? As other objects get farther away, their sizes should also be magnified in the same way.
The looming effect raises the apparent elevation of an object, and globe believers use this to explain how we can see the tops of objects that should be below the horizon. How can the flat earth model reject looming in this case but accept it as the reason that the size of the sun doesn't change? Actually, I don't understand how looming is associated with magnification in the first place.
How does the flat earth model explain the setting sun? It can't be looming or even refraction in general because those have the opposite effect.
You need to realize that "modern scientists" wander off through mathematical models based on assumptions and presuppositions and end up on a mad carnival ride spinning in their mind.
Accurate predictions by a mathematical or physical model affirms the validity of the model. That is science. On the other hand, I agree with you that some physicists can wander into fantasy land with their mathematical models. Some will create a mathematical model in order to generate hypotheses and then end up just presupposing that the model is valid.
I can admit when I've made a mistake, can you?
I certainly can if you can point it out.