Simple experiment really.
Use any tool which can show level. Use an app, use a bubble level, make a contraption with 2 containers of water connected with a tube so the water runs freely.
Now use this tool at a beach with a clear view to the horizon, show that it detects the horizon as being at eye-level.
Then use this tool when at altitude, on a plane, a mountain, a skyscraper, show that the horizon is still at eye-level.
This experiment would show if the horizon is at eye-level as stated by FE, or if it is below eye-level at altitude as stated by RE.
So far, I have yet to see any FEer do this simple experiment EVER, and I have only seen REers do this experiment at altitude (where it shows a dip in the horizon) without showing an accompanying test at low altitude for comparison (though in most cases, this is not required, as they use decent measuring gear like the containers of water, which we know to be level due to physics).
So if a FEer performed that and said they got the FE result, that would be enough for you?
I would hope they at least would take some photos to share.
I generally believe the video results and images shared by almost everyone. I have a harder time using the given claims from them.
Just a recent video on youtube, an FEer claimed he could see a building 22 miles away. He included the calculated hidden height from a curvature calculater (easily verified) and he included the building height that he saw. The problem however was that he completely neglected that the building was built at an altitude, he also didn't factor in that the sea level was really low, and finally he didn't factor in refraction (which makes a huge difference at surface level and especially when taking video over 22 miles of water).
I trust the observations, especially those I can verify due to the location being stated. But I would need all the relevant data aswell.
In my stated experiment, I would at least like the location of the high-altitude experiment, and I would like a clear image of the result (as in a photo taken along the tool to show if the horizon is eye-level or below).
I have an app installed called Dioptra, it can show the amount of degrees to the object I point at. As an app programmer, I know how bad such measurements can be, and I can easily see through experiments that it is 1 degree off constantly. Accounting for this 1 degree (I wish the app could be calibrated properly) I can now do experiments at all altitudes and check the results, though in my whole country, the highest point is 172m and far from any ocean to show level, so I have a hard time doing the experiment myself.
Lets see how an FEer tries the experiment I had:
holding the tool and camera in his hand very unstable, but at least the concept is there, taking a measurement at 2m and 60m.
Then we have the REers test:
the tubes are way more stable when held against the ground, the tubes are wider and allow the water to flow more freely, it also takes measurements at altitudes of 2m, 350m and 1700m, clearly showing a dip in the horizon.
So, the results are clearly opposite, the FE claims no dip, and the RE claims larger dip. Which should I trust? I believe both of their observations to be true.