Coronavirus in rural America (Part VII): ecological measures of rurality in a pandemic

An interesting metric for measuring the concentration of coronavirus incidence came to my attention yesterday when I was studying the New York Times Upshot chart re: the highest concentrations of the pandemic in the United States.  The chart the NYT published is here:

What jumps out at me is that places like Show Low, Arizona and Mount Vernon-Anacortes, Washington are being "measured" based not on the municipality or metropolitan area, but based on the entire county in which the municipality or metropolitan area sits.  So, Show Low, Arizona is actually a municipality of about 10,000 folks, though this table shows the population for all of Navajo County, which covers nearly 10,000 miles.

Similarly, Mt. Vernon-Anacortes, Washington is shown with a metro area of 129,000, but that is actually the population for the entirety of Skagkit County, which stretches far to the east of Mt. Vernon, and even farther to the east of Anacortes.  It's total land area, though, is a fifth of Navajo County, Arizona, at just under 2000 square miles.  My point here is that this isn't really "apples to apples" with New York, New Orleans, or Seattle, all densely populated cities--at last relatively speaking.  (By the way, here is the Los Angeles Times story by Richard Read on the spread in Mt. Vernon-Anacortes, a spread that accelerated with the help of a choir.  I've not read any coverage on the details of the spread in Navajo County, Arizona).

Another interesting example of the apples-to-oranges phenomenon here is Pittsfield, Massachusetts, with a population of less than 50K, the seat of the County of Berkshire, which has a population of about 130K (as suggested by the table), in Western Massachusetts.  The entire county, at 960 square miles, has less than half the land area of Skagkit County, and less than a fifth that of Navajo County, Arizona.  This well illustrates disparities in land area between counties in the Eastern U.S. and the Western U.S.--bearing in mind that Berkshire County, MA covers a great deal of land area as New England counties go.

So, the question arises:  Which is the more appropriate measure--for policy makers considering the situation or in a graphic like this?  The population of the city or municipality where the coronavirus is concentrated, e.g., Show Low, Mt.Vernon-Anacortes?  or the entire county, e.g., Navajo, Skagkit, Berkshire, which is sometimes vastly greater than the city in territory, as shown in this?  And of what salience to this question is the very different situation of New York City, with its five urban boroughs and very high population density?