Muddy, yes, which is why the algorithm would have to be spot-on and the VR set would require a dedicated scaler specifically designed for this task. Of course this is all speculation at this point, so the conversation has very little sense. My point is, when the screen is an inch away from your eyeballs, you want the physical pixels to be as small as humanly possible - upscaling the final image is the least of your worries. It's better to upscale than to see visible gaps between the pixels.
For anyone who wants to see what I mean, take a 256x192 picture, full-screen it on a phone with a high resolution display, put it close to your eye, then display the same image on a DS and do the same. One image will be crap and one will be managable to look at. This is because with a higher physical resolution, you can at least approximate the correct colours. With a small physical resolution, you can do nothing - it's just small and you have to grim and bear it.