Talking about merely the premise of the experiment, it should work. Really, you don't need to come up with predictions. You only need to come up with predictions if you are forming a hypothesis. However, the hypothesis is already formed (BLH). You are merely conducting an experiment to prove or disprove it. Which can be done with only moderately accurate equipment. If there is a change of luminosity when the laser is pointing vertically compared to it pointing horizontally (oriented to the surface of the Earth), then BLH has some support. To further this support, you would need to take accurate measurements at varying angles. However, if the luminosity is the same, then you do not need to collect so much data. Only enough data to clearly show that luminosity is not affected by the orientation of the laser in relation to the Earth.
If there is a variance in the luminosity, then you would need to use BLH to accurately predict the variance at the different angles. BLH would then be put to the test again, to see if the predictions are true. If BLH passes both tests, then it is further on the path to becoming an actual theory, and not a proposed hypothesis. If the FEer's say that this test will not work, then it is up to them to provide a way to properly test their hypothesis. If they can come up with no way to test it, then it is not a hypothesis, and thus, certainly cannot be a theory, as it will no longer fall within the scientific realm at all.