I agree with you on this piont, but I think the reason that the "pointing devices" cannont be used for compeditive gameplay more because the system reads their location/movements from a 2D view.
The assumption that deeply immersive VR requires a true 2-directional brain–computer interface discounts, I think, how good we humans are at voluntarily “remapping” our senses and motor controls to produce various kinesthetic body transfer illusions. It’s possible, even, that mechanical controller devices might provide better control over a VR avatar that a true BCI, as suggested by present day video game players that have tried a wide range of available controls having found that more “realistic” controllers, such as pointing devices, are often inferior to joysticks or even buttons and keyboards, especially in competitive gameplay.
People see in 3D because the image from our left eye overlaps the image from our right eye. The result is being able to see around the side of the object, as well as the front. Modern devices like the X-Box kinect and Nintendo Wii, lack the over lapped vision that we have. They are not able to properly perceive our movements from the 2D veiw point.
Also they have a reaction time that is very noticable. Avatars are not able to move as fast as the players do, so the lagg/ping time make the gameplay seem "off".