> No, they show what the respective government datasets contain
Which matches popular usage in the respective countries. Your example is a non sequitir for explaining why matching popular usage can't possibly work.
For the popular usage of major geographical features to change is extremely rare, so I am skeptical that it's a huge amount of work to keep up with changes in popular usage. People made maps before they had automated pipelines for ingesting geographical names.
My point isn’t so much that it would be impossible to do, more that it’s giving more power to the tech giants to define things however they wish (given they get to define “popular”). In this instance that would be a good thing but as a precedent I’m a lot less sure.