Thanks for reading my rant.
In regards to "Having a cockpit without reference points might become nauseating", yes but you don't actually need to make it a transparent ball in a practical sense. You could just implement it in your current cockpit setup without changing anything to the non-control aspect of the game. There is enough field of view to work with. In the practical sense, it is just you and your cockpit is still while the view through your viewport is changing.
"Dragging a virtual handle to rotate the craft will not result in an immediate reaction".. .This maybe true or not. Not sure without trying out implementation. But I think you maybe over thinking the movement.
Don't think "Aerobics" movement in terms of scale. Think "Gun play". Gun play games don't use a large scale of movement but do a lot and fast. Have you played some VR gun play games. Liked "Pistolwhip"?
Oh when I say "drag" , I don't mean (analogy) a car without power steering. More like moving around a virtual mouse upright on a 3D space. With really big lateral movements, you can use a swipe motion like on laptop trackpads but with your hand. Adding some rumble on the controllers would be nice though.
There have been attempts to map flight game controllers to a mouse but were lackluster. Think it would work better in VR.
"Freelancer" comes to mind...
Would it be difficult to get a barebones test using your current game? I just took a few classes in C+ and Java like a decade ago. I don't really know what is required. Was just wondering.
Funny thought: VR game play developers should take mime lessons. Might help with ideas.
Side question: There are 2 machine gun modes in your game. Is it just one is slower than the other?