Comment Re: "Not Reproduclibe" (Score 3, Insightful) 618
Deducing gravity by noticing what has fallen and how fast is "a bunch of data points and an opinion". It's also falsifiable: you predict what the data will be for other, related circumstances you haven't measured yet. A lot of astronomy, chemistry, biology, and social sciences are done this way because strictly controlled experiments are very difficult.
The ability to make precise predictions, or to give accurate and verifiable _ranges_ for results, seems to be a very good basis for both engineering and science.