You develop for large and small screens, with either keyboard/mouse or touchscreen inputs. That's all.
The specifics are supposed to be automatically handled by preprocessing and deployment tools, or only cared for in the optimization/final polish phase.
Where is the progress in that?