I should mention again that the models I'm using are freely available and easily customizable. Just send me a PM for the link for the google sheets. It might interest you to look with greater detail into Madrid vs NYC vs USA vs Italy, etc, and perhaps model your own test assumptions.
Speaking of NY, they are completely distorting the rest of the US. In fact, US (less NY) daily fatalities peaked a few days ago at 90, and are now at 40+-% of the top:
I've got to give credit to kabar - he's one of the first who brought up the developing awareness about apparent CV specificity. More than anything, it may not be age and/or HCS, but simply dense urban life styles and living conditions that are the prime driver of CV - both scope/rate of infection spread and mortality. This episode is going to launch a thousand PhD theses producing very complex, detailed analyses.
Speaking of NY, they are completely distorting the rest of the US. This is an important observation that can't be repeated enough. Trump and his team will look at the data down to the state and county level to determine how we can return to work on a phased basis. You can still contain and mitigate the virus using good practices at the local level. If the data continue to show a lower and lower mortality rate, then we can adjust our public policy accordingly.