Wait, you're talking about case fatality rates. I thought the above discussion was about the real fatality rate. Obviously CFR is going to be higher because you're excluding all the infections that didn't become cases - which are also the infections least likely to lead to death.