Why do we still have to separate UV channels by mesh on new models?

[AdelQ]

It used to make sense in games like Brawl where they're simply diffuse maps overlayed on each other and can be worked around with no visual drawback. But in games like Breath of the Wild where a normal map will be on a different UV map than the diffuse, it's just counter-intuitive. Why is it still a requirement?

[squall789]

It used to make sense in games like Brawl where they're simply diffuse maps overlayed on each other and can be worked around with no visual drawback. But in games like Breath of the Wild where a normal map will be on a different UV map than the diffuse, it's just counter-intuitive. Why is it still a requirement?

I hope this doesn't count as a necropost.

You're misunderstood a little.
Normal maps aren't usually on a seperate UV map, normal maps share the same UV's as the diffuse texture in the majority of games, a seperate UV map is often used for a detail normal, or used for specific masking effects.
In your example of breath of the wild, I haven't looked at the models myself so I'm guessing here but, the first uv channel is going to be for the normal, diffuse and roughness/metallic/specular texture (I put in all 3 as i can't recall if zelda uses MRAO or standard specular), the second is probably for things like dirt/blood and/or effects.

Lots of games also use them for seemless tiling between objects for example.

For example you might want an effect to go up Links arms in a particular manner, it'll be easier to map your UV's in that way instead of being forced to follow the original UV's.

Multiple UV sets will never go away, they're a real godsend honestly.