Help with constraints and master bone

Hello, I’m new to these forums so sorry if I break any rules. I don’t think i’m violating any policy in this post, but I apologize if I do break any rules. I also apologize if this post is too long and confusing, but I really need some help. So I am trying to re-create the “Wheatley” character from Portal 2. I am trying to make the rig as accurate as possible to the in-game model so it moves identically to how the in-game one moves. I have pretty much nailed most of its functions but still am having troubles getting a few parts to work the way I want them to.

I am having trouble with the constraints on the “bands” as I like to call them. If you are an observant one youmight’ve seen on Wheatley he has these “bands”, here is a picture:

(the bands are hidden on the left and right and top and bottom, the ones looking like they hold the outer “eye” in place, you can also see it on the circular opening on the right of wheatley.)

P S most of the bones are hidden in the pictures.

Now, I want the bands to follow the “eye” of the rig but only on the respective axis that they lie on, as shown in the below picture. I can get this to work if I set the bands to copy rotations of the eye bone on only one certain axis, and the space is set to “Pose Space”. Here’s a picture of it working:

Now the problem is that whenever I rotate the BASE bone, the one that acts as a master bone to move all the armature, or when I rotate any other bone the eye is parented to, the “bands” always tend to get disoriented and rotate on their own not relative to the local space without me rotating them, as shown:

To fix this problem, I just changed the “band” rotation constraint space mode to “Local Space”. Now the master bone can translate itself freely while keeping the other bones in place. But this brings its own set of problems, such as the “bands” not rotating properly when I rotate another bone, one that controls the “eye” bone and another piece of the model although the “eye” bone is parented to it. (In the picture you can see I’m not rotating the “eye” bone but another bone that controls the “eye” bone and another section of the model)

So my two questions are, if you can understand what I’m saying, how do I fix this so that the “bands” can rotate the way they should, along one axis controlled by the “eye” bone, and the “eye” bone can be controlled without interruption by other bones and still have the master bone rotate scale etc. perfectly fine without disorienting anything else? My second question being why does “pose space” and “local space” make such a huge difference?

If anyone can debunk this I would be extremely grateful.

If anyone needs the blend I can supply a version that focuses on the problem

That would help. The problem is parenting, or a world space/local space mix up in the constraints. A blend would help us narrow down the cause.