bartrein,
I learned how to program using an interpreted dialect of FORTRAN called WATFOR in the late Sixties. One of the big advantages of using the interpreted language was the run-time error messages, which could point to the exact line of code that had blown up. I also enjoyed eliminating the job control language needed to compile and link code in a compiled language. In contrast, compiled versions of FORTRAN had informative error messages only when you committed a syntax error. Run-time errors gave you a core dump that was hard to interpret, despite my motivation. As a result, I could write and debug code much faster using the interpreted language than with the compiled equivalent.
The WATFOR interpreter ran on an IBM 360. I imagine the 360 had computing power somewhat akin to an IBM PC AT from the mid Eighties. As a student, I was allowed 60 seconds of computer time a day. I was doing ray tracing to design a mirror to display stars on the ceiling of my dorm room. Despite the inherent inefficiencies of interpreting one statement at a time, it was amazing how much calculation could be accomplished. After fine-tuning the algorithm, I was eventually able to solve the complete problem within the 60 second period. I suspect that the same calculations could run in an interpreter as fast as I can type on a modern PC.
If you are working on a program intended to be used by many people, it might be worth your while to spend more time writing and debugging the code in a compiled language. You gain both execution speed and security (humans can't read machine language very easily) by compiling the code.
In the early Nineties, Microsoft product managers knew that interpreted languages could run fairly fast on a 486 CPU. They probably expected most Excel macros to be used by a single person or perhaps a small group. The typical "developer" would probably be working part-time on the project. It was therefore more beneficial to offer "acceptable speed" while making it easy to write and debug the code. They chose to use a dialect of their very popular Visual Basic language for this purpose. The resulting VBA language was so successful in Excel 4, that it later spread to all the other Office applications.
Using VBA, non-programmers could write very useful code. And they could do so without ever needing to submit a project request to the MIS department and wait for resources to be allocated. If you look at the Microsoft MVPs, I'll wager that very few of them for the Excel/Word/Access applications started out in the MIS department. Most of the ones I have met are gifted amateurs who program on a part-time basis.
Brad