Map is solving the real problem of code duplication for the scenario "create a new list whose elements are the elements of the original list to which f was applied". Not only does it reduce boilerplate but when you read map(f, list) you know that the result is gonna be a new list of the same size, that if f = id you're gonna get the same list, etc. In other words, map has invariants. It captures a tiny subset of all the for loops you can write.
Saying "we don't need map, we have for", is akin to saying "we don't need toUpper, we have for loops" or "we don't need functions, we have goto".