I guess one needs to consider the US is geographically much larger and most land doesn't actually contain people. Considering the density is wiser, but even still. Libraries per occupied area still isn't a good metric. There is no good metric.
What's more important is the qualitative offerings and impact:
1. Spectrum of a. most common services and collections offered everywhere to b. the most comprehensive of those offered by a specific library.
2. What people can do at them: read, research subjects, borrow things, accomplish tasks, host meetings, etc.
This is very hard to measure and not something a business person running the government "like a business" would understand.
IMO the most important metric is "what percentage of kids can walk to a library without asking anybody".
But nowadays people have internet, so I guess it's not THAT important anymore. The ideal library is just a website that lets you download pirated ebooks for free.
The utility of the brick-and-mortar is that some/(many by state) libraries include services and physical items that can be checked out besides media. Plus, besides free Wi-Fi and meeting rooms, it's a non-consumption location to exist in a physical public space. There aren't many more free spaces in America. And, there are millions of people who can't afford internet, a tablet, a computer, or have a place for books. Millions of books and historical local newspapers don't exist in electronic form!
But no, really, (most of) America is truly unwalkable for almost any activity.
To build a library close enough to most American kids that they could walk to it would likely be infeasible, because so many people live places that are simply not walkable at all.
Can I walk a mile? Yes. Can I walk a mile from my house? Not safely.
There's a road I would have to walk down to get to anything that isn't on my street. One lane each way, no shoulders, no sidewalks, and very poor visibility of oncoming traffic due to an inconveniently placed hill.
The nearest public library to my childhood home was 2.2 miles; I would occasionally bike to it before I had a car, but it involved crossing one major street.
What's more important is the qualitative offerings and impact:
1. Spectrum of a. most common services and collections offered everywhere to b. the most comprehensive of those offered by a specific library.
2. What people can do at them: read, research subjects, borrow things, accomplish tasks, host meetings, etc.
This is very hard to measure and not something a business person running the government "like a business" would understand.