I'm not an expert, but I know some about piebaldism-effects.
Interesting markings are often caused by a lack of pigmentation. So, in the way that merles lack pigment and can sometimes be deaf, the same goes for dallies. Pigmentation controls blindness/deafness, but it honestly isn't even very understood by science. With humans, albino individuals have impaired vision, and people with Waardenberg's syndrome have lack of pigmentation, brilliant blue eyes of bieyes, and high instances of deafness.
In dogs that have the extreme whiteness and lack of pigmentation, a double dose of the "Extreme lack of markings" allele can have a similar effect to a double merle. But, the reason people say dogs have to have colored ears is because if the dog lacks pigmented hairs inside the ear, it will be deaf. This happens in all breeds with piebaldism-like characteristics, though, I've seen Siberians that are deaf/blind because of lack of pigmentation. It's just really common in Dallies because most of them lack a lot of pigment, some just have only one copy of the "extrmeme lack of markings" (forgot the name)allele that makes the dog especially white.
ADD: Now I'm really interested, thanks a lot :b
But anyways, it appears that all dallies have exactly the same genetic type for their first two coloring allele pairs, AA and BB. This is interesting because AA is the agouti modifier, and two copies of it prodouces a solid black dog. Even more interesting, the dominant gene for dallies it BB, producing black pigment always. However, they carry a double copy of the sw allele, which is what I was calling the "extreme lack of markings" allele. So, the double sw allele means a completely white dog. But, another gene allows the underlying black pigment to show through on random spots, so it seems all that saves dallies from deafness is luck and a either a double or stronger influence of the ticking, piebald and flecking loci for the black pigment to appear on the coat.