Why are the two modes actually even separate? Third person cameras have a variety of zoom levels, so why isn't the first person camera simply a zoomed in version of the third person? Have the highest zoom be a camera node that sits about an inch or two in front of the avatar's face. Then, when looking around, you would see the third person body and animation taking place below you and behind you.
um
are you for real bro
like have you played Mount & Blade or ARMA? or the first-person mod for GTA4? do you know how clunky movement is? how hard it is to do anything at all? [censored] in ARMA trying to climb a set of stairs is a simulation in itself.
first-person games tend to avoid showing bodies because it makes it a lot easier to track which part of your body is visible to the AI if there are minimal parts to your body. generally if your head is hidden enough, they can't see you. when you get into body awareness you have things like elbows and legs sticking out of cover and if the AI doesn't respond to it then people will call it [censored] AI and if you can't get your arms and legs behind cover properly then people will call it horrible controls.
there's also the matter of legs in first-person - like Crysis, for instance - always looking like you're sliding on ice every time you turn around or move or do anything. they also don't respond to FOV changes realistically, so if you zoom in with your gun or binoculars or something your legs are just as big as they were unzoomed.
in order for body awareness to work well in a game that allows both first- and third-person perspectives, both first- and third-person perspectives need to receive an equal amount of work.
just popping a camera in front of the character's face and calling it a day is like buying a chair to sit on while you eat dinner, and then using that chair as your dinner table while you sit on the floor.