I just watched this Veritasium video, explaining the difficulty of measuring the one-way speed of light. The presenter explains that a round-trip will always average out to $c$ even if the speed in one direction isn't always the same as the speed in a different direction. He used the example of an astronaut on Mars synching his clock with Earth, and how different aberrations would still always make it end up looking like it was consistently $c$.
But the example of Mars got me thinking. Mars has two interesting properties: it's far enough away from Earth that there's a very noticeable time delay, and it's in a different orbit around the sun than we are, so its direction relative to us is constantly changing. So here's an experimental proposal that doesn't require any synchronization: Put a radio beacon on Mars that pulses in a regular, well-defined rhythm, then observe it from Earth for a Martian year or three.
No clock synchronization is necessary; just observe the rhythm. If the speed of light is not constant in all directions, this will create a Doppler effect in the pulses that can't be explained simply by the movement of the two planets. If no such Doppler distortion is detected, we can safely conclude that the speed of light is invariant in different directions. (Or at least in all the directions along the essentially 2-dimensional orbital plane of Sol.)
Is there anything that would cause this experiment to not produce correct results, if it were tried?