I think you are talking about the problem of considering just one person / city.

Big cities can have schools, squares, streets, etc all named after a single person. But they will all be considered as a count of 1. So one city of, say, 2 million has the same "vote" as a small town of 10.000. If that's is what you're thinking about, then yes, you are correct.

What happens in the osm data, though, is that the exact same street can be tagged multiple times, sometimes with slightly different names by different users. That would result in dozen of occurrences for the same person even though there is just one street named after them. The one person/city is a way to avoid that. With the unintended consequences, as you've mentioned, unfortunately.