[intro.execution] Is it possible to clarify the second bullet for immediate subexpressions

As stated in the definition for immediate subexpressions

The immediate subexpressions of an expression E are

• [...]
• any function call that E implicitly invokes

Does it refer to the implicit function call that appears in the lexically part of the expression E or any such function call invoked when evaluating the expression E? I mean, in the following two examples

class Test{
  Test(){}
  ~Test(){}
  int a;
};
int main(){
   Test{}.a; // this expression will implicitly invoke the constructor and destructor of class `Test`
}

In this example, these implicitly invoked function calls appear in the lexically part of the expression Test{}.a.

As a contrast, see the second example

class Test{
  Test(){}
  ~Test(){}
  int a;
};
void fun(){
  Test t{};
}
int main(){
  fun();
}

In this example, evaluate the postfix-expression, namely the function call fun(), will implicitly invoke the constructor and destructor of class T which appear in its function body.

It's unclear what's the intent of the bullet second. These two kinds of understanding have a large difference.

Start with the quoted definition of immediate subexpressions and compare its wording against the two Test and fun() examples in the issue. Clarify whether the second bullet covers only calls lexically within E or also calls made while evaluating E; done when the intended scope is stated unambiguously.