What type of weighing instrument do you propose using? I checked out the Kern (‘gnome’) results again, and the accuracy was not great, probably due to it being an ordinary weighing machine with an ordinary spring. The change in g per 1000 metres of height is about 0.30 Gals, the change per 5 degrees of latitude is 0.45 Gals at 45 degrees latitude, so you need an instrument with accuracy of about 0.1 Gal, otherwise the experiment is meaningless.
For that type of experiment getting reliable readings from a scale put into a vacuum would be the most difficult part, so there would need to be a discovery stage which will determine which type of equipment is most appropriate for a vacuum and pressure changes - as it is possible that calibration can be messed up in that process to produce an unreliable result if they are not designed for that. That's the part I have doubts on. Reaching out to precision scale manufacturers would probably be best for suggestions. It is also possible that there might already be special precision scale devices on the market that have a built-in vacuum chamber, and would be the best case scenario.
Luckily the expected result for UA is a null result (at least under a 'pure-UA' theory), which would makes things easier if there is some sort of unmitigated error somewhere.
But then when you're shown video of the Turning Torso experiment which clearly shows the building being occluded more with distance you invoke "waves" or something.
If you think your Bishop Experiment is a good test of a FE - and I agree it is although I'm sceptical about your results - then some more controlled versions of that would make sense. I believe some FE people had a go at recreating the Bedford Level experiment but I don't think they documented it well.
For a water convexity test, I think by this stage everyone agrees that it is possible to see further than should be possible on an RE. The next step in that sort of test is to do some of the following:
- Take long duration time-lapse photography of the effect and catch how it transitions between obscured and visible scenes, to try to determine which one is an illusion. That might involve leaving a telephoto camera set-up on the shore of a lake for a timespan of a few hours. The risk is that it is possible that only one version is seen, an inconclusive result. But it would provide evidence of limits - how long one version could last if it was an illusion.
It would basically be looking for more transition evidence of this:
https://wiki.tfes.org/Sinking_Ship_Effect_Caused_by_RefractionSomething that clearly shows that a refraction transition event is going on.
- Make a long line of posts like in Rowbotham's
second experiment in ENAG and set up a long duration time-lapse photography device that observes how the top of the poles are aligned for a long period of time. This would provide a set of known point of reference for points along the length of the scene. May take more effort.
I consider the water convexity tests to be "harder" since a single observation taken down at the waterline is no longer sufficient, and the purpose is now to catch something in the act. Since there is an element of refraction, any result for this kind of experiment will still be questioned, since anyone can say that there are multiple layers of refraction effects.
Anyone can say anything, really, and a comprehensive test would require a lot of thought.
For RE evidence that sinking reflects a globe:
- Determine that the sinking consistently occurs to produce something in accordance with the predictions of RE, without needing another theory about "refraction did some of this". That was one of the problems that came up when RE has doubled down on the sinking images in the past.
This sort of investigation relies on an 'analysis', and so it becomes more questionable if the results are not what is predicted by an RE, like what was seen when we had past discussions.