Boris, just to be clear - this is the exact same problem as was seen on the board that you sent back, right?
I have been digging back into this thread and it looks very similar... If that is the case, then it has to be something in your setup.
Your method looks good to me and you dont need to hide or close down everything in your appartment...
There is good distance to electrical stuff, so that is probably not your problem. Also the fact that you can repeat it in the garden points to that not beeing the problem...
Your power cable is quite long and that could possibly be the source of the noise. I would start by cutting that down and twisting it. The battery can sit very close to the board, and you dont need to "copper-clad" it. In theory that can also give eddy-currents in the copper, but that would be a highly unlikely guess....
I normally use a 460 mA 3S and a short (<20cm), twisted wire.....but I have also done calibration with a 2200 3S and not seen adverse affects.
Am I to understand that you are using a car lightbulb as a fuse? That will also add noise together with the long wire.
Otherwise, please try to just let the board sit on a table outside the container and ziploc bag for a while and see if you can reproduce the mag noise. Try and do some error finding by eleminating elements of your process..try another bag, no bag, other container, no container etc...