Ok now this makes sense. I guess it's mostly because even if *we* don't personally care about the "C" genes, the plant does. So while we for example we want the A1 to give us more sweet taste, we still assume the plant will have stem, green leaves and all that stuff coded by these "C" genes. So we just want to change the A part but leave C largely untouched - which is not possible with mutation breeding but possible with CRISPR.