Your comparison to assembly/Python doesn't fit at all. You cannot start in on vector analysis without an understanding of geometry and algebra. You can start Python without understanding assembly at all. You can go back and learn assmebly after you've learned Python but this isn't true in most cases with higher level math. How do I explain the chi-squared distribution without wasting the entire semester on basic maths? Math isn't like an Apple product, it doesn't "just work."
Some make the argument that how we teach STEM based courses can be adjusted so that we start from high-level stuff and teach low-level mechanics along the way. While this approach is intriguing, it will never work. It relies on an assumption about students that isn't true: they are actually interested in the subject. If I had to guess the percentage of students with a deep interest in their subject I would put it at a little bit below the actual graduation rates. Most students are in degree programs nowadays because it is the next step, all of their friends are doing it, it is expected of them, and/or they are trying to make their parents happy. Many of these students will find out what it is they really want to do (whether it is in their first field or their second field), many will recognize that they hate their field and still slog through for social reasons, and many will fail. College is a life changing experience for most people, no matter what category they fall in.