I don't understand, through these examples, why the introduction of a name is necessary in order that an empty return statement be used.

The name doesn't even appear in the function. A name that is declared but not subsequently mentioned serves no purpose (or some side purpose/hack).

In the "NoNamedReturnValue" variant, the return type is still declared. The compiler could, from that type alone, infer that "return" means "return a representative default instance of that type".

That is to say, why can't Go programmers just have this:

  // oi name removed:
  func NamedReturnParams(i int) (/* oi */ objectInfo) {

	if i == 1 {
		// Do one thing
		return // wee, allow this anyway!
	}

	if i == 2 {
		// Do another thing
		return
	}

	if i == 3 {
		// Do one more thing still
		return
	}

	// Normal return
	return
  }

Also, why can't the compiler just optimize away the "return Objectinfo {}" statements down to "return", if those really are equivalent.