In short: if your current assumptions were correct, neither FE nor RE could be correct. In fact, the immediate surroundings of my home couldn't exist. Something is extremely wrong with your suggestions, but without further clarification it's impossible to pinpoint.
You've misunderstood what I'm talking about. Let's try some pictures. Consider a street scene, like the one I described:
Forget about the horizon, forget about any vertical positions, and just look at the apparent width between the successive lampposts. It reduces as they get further away from the observer, even though in reality the posts are equidistant - that, I hope, is a consequence of perspective that we can both agree on - yes? So the azimuth angle subtended at the eyes of our observer, between successive pairs of posts, is getting smaller as the viewing distance increases. Agree?
Now consider a large constellation, such as Orion:
Orion measures roughly 20 degrees tall by 10 degrees wide (note the x-axis on diagrams like this tends to be 'hours' as opposed to degrees, where one hour = 15 degrees). If perspective was playing a role in how we view things close to the horizon then, as it moved from, say, directly above us to being close to the horizon we would expect to see a reduction in size (and indeed a change in shape) of the constellation - agree? But we don't. You can measure Orion anywhere in the sky and, barring minor changes for refraction close to the horizon you will observe the same roughly 10 degree x 20 degree shape.
How does FET explain this anomaly? You can't invoke perspective to explain a phenomenon, such as the stars disappearing below the horizon, if there is no discernible perspective effect on the stars when they are visible as they move around the sky.
[edited due to picture error in first attempt]