And using elevators in the building when they go pick up food in the entrance.
The Worldometers.org website recently added
county-by-county data for certain states, including New York and New Jersey. This makes it much easier to compare areas.
It is certainly true that New York City [which includes five counties -- New York (Manhattan), Kings (Brooklyn), Queens (Queens), Bronx (the Bronx), and Richmond (Staten Island)] is uniquely positioned to see rapid spread of a virus. The mortality rate there is already over 0.25% and it looks like it will end up at or near 0.4%. I don't think we will see rates that high anywhere else in the nation.
Interestingly, the mortality rate is not as high in Manhattan as it is in Queens and the Bronx, where there are fewer elevators and more walk-ups. I guess your point stands if you say "stairwell" instead of "elevator", so I'm not sure what to make of this, except to caution against lumping all of New York City into one bucket. And that applies doubly to Staten Island, where the mortality rate is "only" 0.15%.
Regardless, the City's uniqueness doesn't explain why the virus has been spreading so quickly in Westchester County, Long Island, New Jersey, and Connecticut. The first three of these areas have hit a 0.1% mortality rate, and Connecticut will get there soon.
Not surprisingly, the mortality rate in the first ring of counties surrounding NYC accounts for a large part of this -- Bergen, Essex, Hudson, and Union Counties in New Jersey are all over 0.15%. What is more troubling is that the mortality rate is over 0.1% in the second ring (Middlesex, Somerset, Morris, and Passaic). Because these areas more closely resemble typical suburbs around other US cities, they provide a possible glimpse at how things could develop if precautions aren't taken early enough or maintained long enough.
Please note that all of the data I've used is a mortality rate, not an IFR or even a CFR. A a 0.1% mortality rate means 0.1% of the overall population dies.