Dunbar's number is largely psuedoscience. Dunbar just took the group size of gorillas, scaled up to humans based on brain size, and then published, without bothering with silly little details like actually testing human group sizes.