I have USB wireless headphones and mike for Teamspeak, the game comes through my desktop speakers - I am lucky, there is only me here and the cats are used to it by now - and the headset is not external sound absorbing.
This means I can run the game volume at 100%, Windows sound mixer at 100% and adjust the volume using the desktop speaker volume. I can adjust Teamspeak volume using the volume control on the headset, it is independent of the computer sound system.
Currently the desktop speaker volume is at about 30% for aircraft and 40% for tanks when I am using Teamspeak. If I take it much higher then the mike picks up ingame sounds and these come out to you on Teamspeak. Besides, I know my hearing is crap but 50% is bordering on too loud even for me. 40% is also a good setting for general use for me.
I see a simple analogy for my setup as 'fire pump #1' (game sound volume) pumping 'water' (game volume and effects) to 'fire pump #2' (Windows volume mixer) and then to the 'fireman's branch' (speaker volume switch) at the end of the 'hoseline' (speaker wires). By having all the 'fire pump valves' (volume controls) fully open on both 'pumps' you are getting all the volume and effects all the way to the 'branch' at the end - I can reduce it here to get the desired volume. To me, introducing restrictions along the line ie. reducing output on one of the volume controls, or 'pumps', creates variables so you could be fiddling around to get the right effect. Just me and my thinking.....