C++ programmer here.
In my team I've actively pushed not to have for loops when something from the algorithms library will do (transform, copy_if, find_if, etc). The reason: Because I usually cannot quickly grasp what the for loop intends to compute - I'm quite surprised you're saying the opposite.
I met with some resistance, as people aren't used to thinking in terms of algorithms. So I was slowly browsing our code base, finding for loops, and converting the trivial ones to calls to the algorithms library. Then I came across one for loop that was, frankly, a pointless loop. It didn't seem to do anything useful. I was sure it was a bug, but I couldn't tell from the context what the code was supposed to do. I emailed the committer, and he fixed it. Then I took the fixed version and converted it to an algorithms call.
Then I made a presentation to the team showing some of my examples, including the one with the bug. That bug convinced many to start using other constructs. Every time you write a for loop for which there is an algorithm in the library, you are reinventing the wheel, and have a good chance of introducing a bug.
(Sorry, probably a bit of a tangent to the whole thread).