That’s an interesting way to present the data.

First thing popped into my head is how railways have slope limit both uphill and downhill. To traverse a slope will require way more rails to loop and/or zigzags. So rails in mountainous region will have significantly more miles to cover the same distance between two points than in a flat region. I wonder if there’s enough data to change the measure from plain miles to actual distance covered. That would make an interesting comparison.