Would you agree that science can help humans make much better guesses (about things for which we have not yet figured out how to do a measurement/observation over) based upon the data collected through the methodology? You’re surely not trying to limit scientists to those returns from measurement and observation. I mean, quantum mechanics is based upon the notion that there are ‘things’ unobservable which only statistical probabilities can make room for in the calculations!
Certainly. That happens a lot. New technology will often produce ways of observing and measuring phenomona that simply were't possible before and existing theories compared to new information when it becomes available. Those theories are developed based on information we know, and a measure of a theories validity is it's ability to predict what future information will show when we have the tools to get it.