Stats, like wind in miles-per-hour? That seems arbitrary. Why not just use wind speed in plank-lengths per zeptosecond, then?

It's almost as if expressive abstractions are helpful for human understanding of large sets of numerical statistics, no?