As I understand it, scientific theories are models based on multiple tested hypotheses. They make sense of multiple lines of evidence. Each hypothesis is testable and makes predictions, but a theory is more overarching than a single hypothesis.
@jammycakes made some good points about observability, testability, and repeatability in relation to models that describe the past here: The Big Bang Idea - #2 by jammycakes
Huge areas of scientific investigation involve trying to understand how we got where we are. These models are testable and falsifiable, but they are tested and falsified in different ways than, say, does this medicine slow cancer growth in cells?
I agree that much of what falls under evolutionary psychology at the popular level amounts to a Just So story. Psychology, sociology, anthropology, archaeology, and linguisitcs (which all come into play when studying the human development of religious practices) are “soft sciences” and don’t follow the same methodologies as hard sciences. Observations in these fields are more subjective to begin with and rarely involves purely empirical measurements. The observations themselves require interpretation of situations and behavior that are colored by the lenses of the researchers and shaped by the investigative tools.