Our hyperfixation with "Progress" has limited science to the production of models sufficient for engineers to invent, design, and build. What most modern people call science (i.e. observational analysis and modeling) is not actually SCIENCE nor THE SCIENTIFIC METHOD. It is but a single step in the scientific method. To completely apply the method, a true scientist must take such observations (and subsequent modeled analysis) and THEN derive falsifiable claims based on those observations.

That last step of testing falsifiable claims derived from observational analysis is what makes science science. It is the chipping away of untruth. The true scientist doesn't limit themselves to "observe, model, and describe"; He takes such descriptions, formulates falsifiable theories and TESTS them rigorously. He tries to break them. A true scientist must be the wildest skeptic on earth.

A true scientist doesn't limit themselves to the question of "does my model fit" (that is the realm of the mathematician and the launchpad of the engineer), but rather a scientist follows up with "WHY does my model fit when it fits and WHY does it not fit when it doesn't". He doesn't dismiss outliers. For the true scientist, the outliers are the most important thing to explore.

Engineers, not scientists, say "this model is good enough". They are not primarily concerned with developing fundamental theories about the nature of the world, but instead with applying these theories in a practical manner. Models are important and valuable, and they may be all the engineer needs. The engineer is perfectly ok w/ outliers and "a good enough fit".