You are, of course correct, but off point. The issue in the original point was unit count vs revenue, not high vs low risk. Given a the choice between a company with A) low unit sale but high ROI or B) high unit sale but low ROI the investor will always choose A. Unit count is irrelevant UNLESS it impact profits, which is clearly NOT the case with Apple. It could impact profits if it was so low as to discourage development for the platform - again not the case here.