You’ve misunderstood. Yes, the 2 clocks show a different time off by microseconds after awhile. First calibrate them when they are nearby so the drift is <1ns/s. Simply by adding 0.1 or 0.2 ns every second. You are writing the 2 numbers down and sending this paper back in the US mail. The distances are the same in each direction. The 2 clocks have not moved. There are only 2 so of course the distances are the same! Let me try again. At the home point write down the time A when you send the signal. Miles away, write down B when you receive it. Write down C when you send it back. B=C if you use a mirror. D when you receive it. Subtract B-A the next day. Subtract D-C. These might be the same within a nanosecond?
It’s not needed to enclose the axis. This is a different experiment as there is some 2D area enclosed. It is a loop and easy to verify IRL. Mine has no area. As it is a line.
No, I don’t understand what you gain from this additional leg. Are you trying to synchronize the clocks somehow? Why? They don’t need to be synchronized.
How did the clocks get to those positions?
They clearly have moved at non-zero velocity from your central measurement point to their final locations, i.e. their calibration (even with a known drift) will be affected by how fast you move them to those positions. If their speeds getting to those positions differs in the least (and how would you know this without a two-way measurement?!), your calibration (even with a drift) is already ruined.
Worse, you are moving them to different positions. So you are assuming the speed of light is isotropic (the same in both those directions) which is what you are trying to establish (or dis-establish) in the first place!
Again, this experiment relies on assuming that a one-way speed of light measurement is the same as a two-way speed of light measurement. It assumes what it is trying to prove/disprove.
No. You move them slow or fast, it doesn’t matter. I agree that moving them will put them even more out of sync, due to Special Relativity. Perhaps even milliseconds. But their rate of ticking will still match. Why wouldn’t it? Since they are now stationary again. I am absolutely not assuming speed is isotropic. Their velocity is the same rotating around the Earth’s axis. Their acceleration is the same to match gravity. Why would their resting position determine the speed of light at all, within the same inertial frame of reference. Are you denying Einstein? I am not making that assumption.
You sync the clocks before hand with a split beam. Then send one beam that does the round trip and a straight trip of the same length as the round trip at the same time.
Rate of ticking as measured from what frame of reference?
This is quite important - the measured “rate of ticking” will vary depending on the relative velocity of the observer from the clocks. So if you deliberately move the clocks from the observer to their final positions, the ‘rates’ each clock tick change with the velocity of each clock along their journey. And how much that rates changes and how much difference it makes depends on the speed of light - the very quantity you are trying to measure.
If you assume the tick rates still match, that means you are assuming the one-way velocity is the same as the two-way velocity. That defeats the purpose of the experiment - you are assuming the very thing you are trying to measure.
So, yes, it does matter.
As AntiBob said: You cannot sync the clocks when they are apart, or together, when you move them later.
You split the beam so you can sync them in two different positions. You do not move them after that.
You’d have to assume the speed of light is isotropic. Which is what we’re trying to prove, or measure. There’s no point in measuring it, if the experiment assumes it to begin with.
That is why you then use a single beam to do the round trip and straight line. A portion of the beam does the round trip, a portion does the one way trip that is twice as long.
Disagree. Time is the same everywhere within the SAME inertial frame of reference. Read Einstein. Inertial frame of reference means the same “global” velocity and accelleration. Virtually none when they are the same for everything in this reference frame. There is some truth, you cannot know the time is the same because you’d have to move or communicate to verify this. Here’s an experiment they did decades ago. They sent an atomic clock on a commercial jet around the world. It came back slower than the one on the ground. Then did the same with the other one. They are back in sync. You can send both at the speed of light to the moon. They would match. If we had 2 moons opposite of each other you can send one to each. They would match only if light were isotropic. Bring them home. Then do another experiment leaving them up for 10 years. This time the moon is fixed not orbiting, geosynchronous. Bring them home. It doesn’t matter where they are sitting for 10 years, they would stay in sync as well as the first experiment. Did that make sense?
Back in ‘sync’ within the bounds of experimental error. Not precisely the same thing.
By bringing them home, you’re back to a two-way measurement again for each of them.
If you mail the measurement from the moon or wherever back to ‘home’, you have to make assumptions about the clock drift from your clock at home vs whatever is being measured and sent back.
I agree you cannot know how far apart they are (time), WHILE THEY ARE APART. Unless you assume isotropic speed. But their gradually drifting apart over years time would be the same, within the same inertial frame, regardless of where they are sitting.
I’m trying to follow. The RT and OW are the same distance, since the OW is double. But we already know the exact speed of the RT, EVEN IF LIGHT IS NOT ISOTROPIC. So how does doing that help? They will arrive at exactly the same time, no matter how you measure this. No wait, they will not. Only if it is isotropic.
The RT spends half the distance going the opposite direction of the OW. If it is slower or faster during that leg the timing should show it. There is the issue of syncing the clocks with a split beam. One half of that beam may go slower, so the sync will not be identical. But it might in fact make a difference more obvious. I have drawn it out and assigned ludicrous possible differences in speed to the legs. I think it might work.
But getting the positioning so exact? Might be tough. A very long distance might make up for minute difference.
Interesting idea. Please keep thinking about this. BUT there is no need to do the RT part. We already know the velocity and time elapsed. It is C. Even if light is not isotropic.
All you have to do is follow my kid sister up the road for about a quarter mile and you’ll have your answer. 
As a bystander to this conversation with nothing much useful to contribute, I would like to ask:
What becomes possible if you can measure the one-way speed of light, that wasn’t possible if you had to measure the two-way speed? What can you do with this capability if you can actually do it?
As far as I’m understanding, it’s more or less academic in that, for all we know, the speed of light might have a universally preferred max velocity. Maybe it’s almost instantaneous in one direction, but half-C in the other, and all other vectors are a blend between the two? Extremely, extremely unlikely, but hey… we’ve never verified this experimentally.