The problem with physics today is rationalism - constructing theories with only partial reference to experimental data. Some fields, like psychology, have the opposite problem - empiricism - where they accumulate countless statistical correlations but fail to integrate them into consistent theories.
I suggest anyone who's seriously interested in a solution to these issues to listen to the following, then look up Leonard Peikoff and David Harriman's other writings on the subject.
https://campus.aynrand.org/campus-courses/understanding-obje...