Underlying principles are nice for science, whatever works is nice for engineering. There is plenty of historical precedent where we build stuff that works without knowing exactly why it works.