thanx for your kind support.
The reason for “group panning” (at a later stage) is, to send a couple of sound tracks to a group (to get one common sound) and then pan the whole group (using automation). Which means, that I won’t need to pan each single sound source around with the same amount of several automation tracks.
I know about the possibility of copying automation data but it doesn’t help me if I want to change the 3D sound position later on. Then I’ve to modify each single automation track of each single source track. Same is related for creating first a “wave file” to have the sound and then panning this new file around just with 1 panner. It might be no usual working case for Nuendo/Cubase but my work is also not usual… 
Ahh, maybe I should keep the source as stereo, then send it to the group and change it there to ambisonics. Should give it a try.
Indeed, that’s the way how it goes!
Thanx for helping me to open my eyes!