Your story about demonstrating it to a car club is confirmation for me that ADS can default to firm with no DTCs then.
I've considered it being a silent failure of the computer. However, I find it hard to believe that some little thing in the computer has failed without it either causing a DTC (since there is a DTC for the computer itself) or just failing to communicate with an SDS. Thus I've diagnosed my problems on the assumption that the computer is working fine.
So it might be worth it to check that the sensors are all firmly placed on their mounts. Maybe the computer is thinking something is happening and stiffing appropriately. If it isn't the computer, then it's data going into the computer.
Which brings me to this little observation. On my car while it was idling, my body acceleration sensor (B24) was bouncing between 2.50V and 2.57V while the wheel acceleration sensor was at 2.50V steady, which I thought was weird. I never have though because it was such a small amount of voltage that I figured it wouldn't matter. However, I can't count on such assumptions when the voltage ranges aren't publicly known. Might be something I should look into. Your system on the other hand has three body sensors, but minus the wheel sensor; as long as it's ADS 2.
As for the work on the headlights, the particular accelerometer (B24/3) for the passenger side is closer towards the coffin box. So yes there is one on that side, but not all that close to the headlight.
I Wish I could take my car to Victor from restoreyourmercedes.com. I bet he could fix it.