I always thought that was what all Physical models were, a best approximation, given the empirical data available at the time.