I think you have the idea backwards a bit. The system isn't interested in making the background blurry, it wants to make your exact eye point the section of the

I think you have the idea backwards a bit. The system isn't interested in making the background blurry, it wants to make your exact eye point the section of the image with the highest fidelity. Then the renderer can just sort of half ass the rest of the image. It's a way to make the apparent image better than what the hardware would allow.