I think that you couldn't be more wrong - at least at the point where kids get taught physics as a separate class (generally at the high school level). Especially since you seem to think all computers can do is simulations. There are three main types of learners - folks who need to see stuff, folks who need to hear stuff, and folks that need to do stuff in order to really grasp a topic. Having computers in your classrooms can really help visual and kinesthetic learners.
I have a hard time visualizing 3D math based problems. I think that if some of the SAS software that is available today had been available when I was taking calculus I wouldn't have had nearly the problems imaging rotating formulas around an axis. About the only physics I was able to grasp were ones where we had done real life experiments like putting together race cars as a team and then using computers with peripherals that tracked velocity and speed and such. Even in less math based classes like chemistry we used temperature probes that plugged into computers and then we did all sorts of manipulation of the data. I got then when an exothermic reaction was. I might not have been using exactly the same equipment I would a year or two later in college but I have to say I was still ahead of the curve and wasn't at all intimidated by the equipment we did use. Then a couple of years later on the job the same was true again. (And some of the equipment hasn't changed that much, for instance pH meters and DO probes are still fairly similar to how they were 15 years ago or so when I used them in class.) When it came time to looking at an excel spread with thousands of data points that needed to be manipulated, I was finishing early and helping classmates in my physics labs. Granted I was lucky enough to be at a great high school (shout out here to all the Roanoke Valley Govenerds - I know there must be some of you on here) where the computers were truly integrated into the lessons for data gathering, manipulations, etc., but learning theory through experience can be really important for some types of learners - like me.
With how the fields of science and math are now, what you are saying is analogous to telling computer and programming teachers that they should teach theory only until programmers try to get jobs because the in vogue programming language of the day changes so quickly and those pesky computers are just going to get in the way of learning the theory. I'm not a programmer but even so I can only imagine what my boss would have said if I told him I hadn't used any of the software before coming to work here (I suspect I wouldn't have gotten hired and that he wouldn't be my boss right now).