You're confusing things. Not everyone who has the disease is diagnosed or treated in a hospital. In fact the majority of people are probably not. As the healthcare system is stressed, you will see higher "mortality" in the hospital cases because only the extremely severe patients will seek treatment and the rest will try to get by on their own. None of this is _population_ mortality. Meaning, these numbers cannot be extrapolated to the population as a whole, due to the selection bias inherent in them, and the lack of accurate infection rate estimates.
This ignores several things. For one thing we might get a reprieve due to seasonality. For another, better treatments will almost certainly be found in the near future. Some are undergoing trials as we speak. Then there's the issue of people who will acquire immunity and therefore cease to be transmission vectors. And then there's the most glaring issue: less than 6K people are likely to die in the _very epicenter_ of the epidemic: in China. Out of 55M people in that region. People are returning to work there now. So this can be contained, with drastic measures. And you can bet drastic measures will be taken if we even begin to approach seven figure casualty counts. Extrapolated from 55M Hubei population to 331M of US population, this would be less than 40K dead: on the order of a severe flu season, but compressed into 3 months. That's in the absence of any advanced treatments, and with a healthcare system arguably inferior to what you'd find in the West.
