I think that we need to keep to the established labels for flash groups - it avoids errors.
FEC applies only to E-TTL.
In the camera menu, a global FEC can be applied,
If in the Wireless menu, the firing group as set to A:B C, then the user is using three groups of one-or more flashes.
A and B are for Main and Fill; C is for background/rim. All three are test-fired in turn. How does the evaluation work if there is no, say, A flash? The camera and the triggers must act as if there is a Group A flash.
Assuming there are three flashes, one in each group, Then A and B together with Ambient are calculated to produce a standard 18% grey exposure. If the global FEC is not zero, then that 18% grey will be reduced or increased accordingly for the two flashes. (The ambient is of course controlled by EC.)
Now the C level can be calculated using its group FEC.
In addition, when using the 622Cs, any FEC set on an individual flash will be added to the total flash lighting.
My main concern is that with only two flashes, rational results will come from A+B+C Global/All, or A:B. NOT A:B C.
Overall contribution of A and B are determined by a normal exposure plus the global FEC. The contribution of A is determined as a ratio compared to B. That is determined by the photographer for artistic effect.
BTW - this is all Canon flash lighting required knowledge.

