When designing a turbomachine, the best analytical approach is not necessarily the most accurate method possible, but the one that will generate the best candidate design in the time and budget available.  It makes little sense to do analytical work down to a 1/10% accuracy,  while ignoring candidate designs that might yield improvements 20 times that amount if they were considered in the first place.  Mastering the art of modern engineering requires an intuitive sense for the cost and accuracy of a given analytical method, and knowing when to move between them.