C++ assignment by value, what are the drawbacks?

Question

Recently, I have stumbled over this pattern:

class Foo {
   Foo() = default;
   Foo(const Foo &) = default;
   Foo(Foo &&) = default;
   Foo & operator=(Foo rhs) {
      swap(rhs);
      return *this;
   }
};

A close look at this pattern revealed that, thanks to copy elision, this single assignment function gives for free both move assignment and the copy assignment by "copy and swap". Since it is so handy, why isn't it a "mainstream pattern". Is there any by doing it this way?

Note: a custom made copy constructor without temporaries could yield the operation in fewer steps but, For this question, what benefit explicitly writing copy and move assignment around swap could bring.

In respect to possible duplicate of question "what is the copy-and-swap idiom", my question assumes the reader already knows what it is.

Thanks