Cool. I will have a look at the log later.
KR is often dependent on the KR sensitivity settings in the tables, so mechanical noise can indeed casue this.
Tomorrow at the dyno, it is a good opportunity to run a complete log and cover all of the parameters. If the dyno shop is using external professional Knock sensors, I.E. Plex
https://www.plex-tuning.com/product/knock-detection/ then you could find out if it is real or fake.
The OEM sensors are not quite as accurate as Plex.
Will you have your remote tuner standing by in the background, or even monitoring your session at the dyno?
This could prove effective because he will able to calibrate in RT while seeing the results.
BtW, I used to be (and probably still am in a way..) a Turbo guy myself - used to have a small 2.0L / 4 Cyl. VW-Audi 20VT platform. My other car today (daily) is also a turbo - BMW with a B48 2.0L 4 Cyl. engine, making relatively high power for this platform with a few supporting bolt-ons and an effective calibration.
Good luck at the dyno and please post the results and let us know details how it went and the conclusions and plans moving forward.