Thank you!

Looking at your screenshots, the circuitry is working as intended.

The motor acts as a generator when rotated by hand. As this is a closed system, the generated energy manifests as a voltage rise which disconnects the PSU and the generated energy powers the system until it is expended and the PSU re-attaches.

Voltage is allowed to rise unimpeded (except for energy harvesting) till it reaches 41V when the circuitry kicks in and gradually bleeds the excess energy off to prevent a hard-lockout at 44.5V. In your case, that would be a voltage rise of 6.35V.

It would seem that your 5V rises are due to the amount of energy generated during specific high velocity counter-steering events taking longer to dissipate than the usual 3.2V type events.

Keep an eye on the disconnects though. It sounds like a cabling related issue causing intermittent USB bus disturbances.

I observed the behaviour of the base throughout several days, managed to get around 5,20 V only once, the rest of the time it's around 3,20 V. No FFB problems at all. I have my torque turned up to 20 Nm and in AC the gain's at 45%. After increasing the gain I actually never got it above 5 V.

UPD: 15 minutes after posting this I now have my base disconnecting every time a high torque amount is reached. I can even recreate the issue when applying high torque from End Stops.
UPD 2: The disconnecting issue wasn't tied with high torque apparently, but with any torque, it stopped by itself after some time, oddly I didn't get USB resets in telemetry, but could see SW disappearing and instantly reconnecting in devices. Looked like an EMI issue, but I never got it before on the base itself.

Down below screenshots of the 2 sessions I had one after another.

Sounds like a one-off spike in an extreme situation. I'd say still not cause for worry but if you could post a screenshot of your Diagnostic tab, we could establish a baseline for your Telemetry readings

I think it was just some sort of spike as it was not constantly running with such voltage (I have diagnostics tab always enabled).

I should also mention that I didn't experience any change in FFB, it was working as expected.

Now I am torn on what to do next. Should I just wait and observe and send the controller box back when it'll get worse? Will it ever get worse or no? What am I risking, controller box damage or both controller box and motor damage? Or maybe it's just an anomaly that I can ignore?

Most of the time the observed voltage rise is around 3,09 - 3,22 V.

4.25V while on the higher side is still within the control range for a SW20E. Nothing to worry about. Over 5V, we would start to investigate why.

The upper limit for the SW30 would be around 14V