Indeed, it's though all the false starts that one gathers subject knowledge. All the people who build "untestable models" for their PhDs have at least learned s

Indeed, it's though all the false starts that one gathers subject knowledge. All the people who build "untestable models" for their PhDs have at least learned something about atmospheric modelling, what numerical approaches work and what approaches work less well.