The way i see it is that you’re effectively rendering at a lower resolution, as you’re cropping your screen, this allows you to run the middle of the screen at maximum resolution and keep high frame rate, doing 0.7x fov means that the gpu has to process half the amount of pixels. For some reason that i don’t understand i can decrease fov to 0.7x and not notice the fov reduction much, going lower than 0.7x vertical would make the square cropped borders noticeable.
Besides the resolution issue, i guess that this could lead to a flatter image on each eye (as in less “fish eye” effect), and the scene to contain less stuff independent to the resolution, however i don’t really notice a difference.