That's assuming everyone's getting coverage, which isn't really the case. What I would really want to see is the density distribution of the parts of the count

That's assuming everyone's getting coverage, which isn't really the case. What I would really want to see is the density distribution of the parts of the country that actually have coverage. Eg how many square miles do I need to cover to serve 50% of the broadband users, 75%, 99%. That would give us a way better idea. Lastly, I wouldn't discount the theory that the profit margins might be bad because the companies are inefficient. When there's no competitor or threat to the business it's easy to be inefficient and still end up with profits.