Your analysis is basically correct. These syntactic features do not play nicely with pointfree style. But I can go into a bit more detail.
It isn't surprising that do-notation prevents point-free style. The purpose of do-notation is to provide a nicer way to bind variables in monadic code. The goal of pointfree style is to get rid of unneeded variable bindings. In fact, if you have a function written with do-notation that can be translated into pointfree style, then you don't need do-notation. For example, beginners will often write code like this:
do { x <- m ; f x }
When they could write the much simpler:
m >>= f
Record notation is a different situation. The notation itself is very pointy, but the notation is optional. Using lenses to access and update your records might be a good way to integrate them into pointfree code.
-Karl