Will make further post re disconnected wheels soon but not now. In simplest scenario, axle is massive so that wheels would be called flanges at each end, looking much like a drum. Each flange has one aperture, as small as possible. One flange has a mechanical adjustment for exact angular position of aperture wrt "axle". Photon source is near one flange, to be decided if attached or not. Photon detector is near other flange, definitely not attached. Operation is go/no go on photon detection. No timing, just a photon strike or none. Device is held at constant RPM by whatever means is best. Minutely accurate tachometer observes and records RPM. That is all the instrumentation needed. We do not need time of entry into first aperture. We do not need time of exit of "out" aperture. We know that the photon has traversed a known distance which is measured mechanically. We do not need minute accuracy in the distance measurement. The distance is the same for all executions of the experiment, and all we want to accomplish is to turn the apparatus in various directions and and see if there is a different outcome of the experiments. We do not need to calculate anything. Each run of the experiment sees us having changed the location of one aperture through several runs, while turning the same RPM, until we get photons traversing through the "out" aperture rather than splatting against the solid part of the flange. We carefully note the angle of the "out" aperture by mechanical measurement. Then we turn the device to a different direction and repeat the drill of adjusting the location of one aperture until we get passage of photons. We have needed no instrumentation at all except for go/no go photon detection, and monitoring of RPM. As experiments are conducted it will become apparent if RPM can be relied on to be perfectly consistent: if that proves problematical then both photon strikes and RPM should be timed so that each strike can be related to a specific RPM. The timing device can be integral with the photon detector and the tachometer in one small device so that synchronization matters are negligible. Comparing the positions of the adjustable aperture reveals to us whether there has been a difference of photon transit time or not. Faster or slower than the right transit time will see a photon splat against the flange beside the aperture. We need not know what the transit time was nor what the photon speed was. It will be sufficient to recognize a difference or lack of one by noting if the same aperture location works or does not work for different apparatus orientations. Of course it will be immediately apparent that the aperture location could be held unchanged while RPM could be varied to cause photon go/no go. Execution of experiments should quickly show which method might be preferable in actual practice. The smaller the apertures can be, the lower the RPM can be. Candidate photon energies and their corresponding practical go/no go aperture size could be an important design factor.