Hi Robin,
That is an excellent question! It does mess with the scale of image textures, but when different parts use different textures, you can ‘compensate’ for that difference in scale. And when you are painting the model, that ‘compensation’ happens automatically.
The advantage is, that important parts can get a higher resolution than unimportant parts. For instance, when unwrapping a human head, you’ll need more resolution in the face, than on the back of the head.
This is by no means a perfect or complete answer, but I hope it gives you an idea of how different situations can require different methods.
