> Until the middle of the 20th Century, cartography was more an art than a science.
Strange perspective, accurate nautical charts were critical infrastructure in the age of sail, making them and using them was a very technical endeavor backed by significant financing. The brightest scientists of the period spent a lot energy on the longitude problem and similar navigation bottlenecks. Accurate land maps were also important for military and state finance purposes. Much of early mathematics and astronomy were focused on measuring the Earth and pinpointing your location.
I’m not sure if map-making was a science explicitly, there was no UX academia like there is now, but it was certainly a serious engineering field rather than an “art”, and dismissing that rich history of know-how seems like a poor foundation for a review of map-making theory.