Courage - I have not analyzed this and asked what exactly makes the differences between various big cities and personally wouldn't feel competent to. But I imagine local lifestyle and culture make a difference, too, and that would not show up in those numbers. I've never been to Hubei but lived in London for a few years in the past and it's quite different from NYC, not only in density that's obvious even when you just walk around on the streets or use public transportation during a normal time, but also in culture and people's behavior. People tend to have a more "keeping a distance" mentality in London (probably in the UK in general vs American culture at least). There aren't so many large apartment buildings, especially with shared facilities. I think people in NYC tend to be quite impatient and live a more fast-paced life compared to everywhere else I've lived, including even London.
I'm talking about London because that's what I know quite well (although it's been >10 years now), probably there are many cultural differences in Hubei and other places in Asia, likely even more. These things don't necessarily influence death rate per se, but I am sure they affect infection rate. Plus there is also what you mentioned, that we just have no idea how accurate those known COVID case numbers are in the various places.
As far as testing, I can speak for my own experience that remains uncertain. When I was sick recently, I really tried to look for an opportunity to get tested without going to the ER or an Urgent Care facility, to no avail so far, hospital even messed up an appointment I could actually get. So no testing. Maybe at some point, if they call me for the antibody testing, but I am honestly less enthusiastic about it now and probably would not pursue very actively, unless there was an easy opportunity. Just like it is being often said, probably thousands of non-severe cases are missed in this way everywhere. Heck, they even miss quite a lot of COVID deaths when they cannot be linked to COVID formally... not sure how many are even analyzed and diagnosed post-mortem.
I understand the curiosity about the numbers and trying to figure something out though. I did a bit similar about the actual biology part that some research groups have published last week (using methods that I also use in my own work so am familiar with), it can be interesting. There have also been countless similar, very active discussions about the numbers and everything COVID among professional and amateur scientists on social media that I kinda followed for a while in the beginning. I personally think, in part, trying to dive into it this way can reflect a need for control via attempting to understand. Certainly more fun for me than shopping sprees and hoarding toilet paper
