Take the data. Analyze the data. I've even done that for you with the original data set. There is zero evidence at all that the mount caused any problems. If you looked at the data itself you'd see. I'm not the only person that has poured over this data. Many people have and all have concluded that the mount performed without a hitch. The CSV cannot tell you what the cause of any of it's numbers are. The data is the only way you can determine that. That's why your take was sketchy to begin with and now it's 100% clear you looked at a CSV and drew a false conclusion from it. No matter whether it is an accurate or inaccurate CSV, it cannot tell you the cause. You literally made up what you thought happened and stuck with that.
That's not how any of this works.
You are incorrect. I didnt make up anything. I drew my conclusion from this premise:
Collimation does not change the variation of eccentricities.
Think of it like this: imagine i set the scope to a fixed field of stars (stars don’t move). I take an imagine and measure eccentricities. The next image is identical to the first and therefore also the eccentricity measurements. Any collimation error is fixed but there is no spread in values.
In your CSV eccentricies go from 0.42 to 0.57 iirc, thats too much variation; I would think the mount with normal loading has much tighter variation and always sub 0.45.
Some alternative hypothesis are:
Scope flex, shifting mirror, loose mirror somewhere, focuser sag, etc but given your equipment I think those are rare.