Looks like the KR is very replicatable. 6.5, wow, with just a short trip into WOT territory. As SHOnUup said, worth a look undercar to make sure everything's secured properly. Could be false knock caused by resonance from a loose heatshield.
Seeing wide variances between AFR commanded vs measured (AFRc vs AFRm), up to almost 16.x, more often up to 4. Assuming AFRm is accurate, the engine looks like it is running lean when the high KR values are obtained. AFRm drops out during sudden WOT acceleration, but that could be a torque/BT thing also, or even normal for the ECU. Can you datalog with the engine warmed up and idling? Also set the Torque data collection intervals to 0.1 second (if you have a highspeed BT adapter).
So I am inclined to believe that there is unmetered air entering the system, meaning loose piping somewhere. Could be a bad MAP or MAP wiring/connection, easy to check with the cover off. Also reset the plug gap to 0.035 and datalog again.
HTH.