Seems like a good idea. I guess the invisible light provides high-resolution luminance data, and the low-resolution visible shot provides the chroma. Since our eyes are more sensitive to luma than chroma, the image is not degraded much compared to a "correct" exposure.