There are a few ways this is advantageous. Modern camera sensors have a bayer filter in front of them - a RGB grid basically. This allows each pixel to see only

There are a few ways this is advantageous. Modern camera sensors have a bayer filter in front of them - a RGB grid basically. This allows each pixel to see only one color of light. The resulting image is "demosaiced" with an algorithm, where by each pixel determines the two colors it's missing by borrowing from nearby pixels. This can cause some artifacts. By squeezing more pixels into the same space, effectively each resulting pixels in the ~8MP image has actual RGB data, not just data borrowed from neighbors.