I had that thought, too. I think it's possible if the resulting baby is missing the normal/functional copy of the same gene that produces the cinnie mutation. So, one copy of the cinnie gene, and the other copy, which should be normal, is either not there, or won't work, so the snake is only left with a working cinnamon gene to tell it what appearance to produce.
Cooking dinner right now, so that's probably a terrible explanation. But one of the hypotheses for some paradoxes is quite similar -- just in the case of such paradoxes, the normal gene is only missing or defunct in sections of the snake, where as this baby is the whole body.