The wording in the parenthesis of the second bullet looks like we don't consider these conversion functions of return type “cv2 T2” and “rvalue reference to cv2 T2” as candidate functions when the reference be initialized is an lvalue reference to a const-qualified object type.

Yes, because that's not direct reference binding. (A temporary will be materialized.)

The wording in the parenthesis of the second bullet looks like we don't consider these conversion functions of return type “cv2 T2” and “rvalue reference to cv2 T2” as candidate functions when the reference be initialized is an lvalue reference to a const-qualified object type.

Yes, because that's not direct reference binding. (A temporary will be materialized.)

Doesn't temporary materialization conversion apply to when the conversion function's return type is cv T and the reference be initialized is an rvalue reference? why exclude lvalue reference to const-qualified type?

As said in: If the converted initializer is a prvalue, its type T4 is adjusted to type “cv1 T4” ([conv.qual]) and the temporary materialization conversion is applied. In any case, the reference is bound to the resulting glvalue (or to an appropriate base class subobject).

Or, Is that just the intent of the standard to avoid to view the binding lvalue reference to a const-qualified type to an rvalue as directly binding? which should undergo dcl.init.ref#5.4.1 or dcl.init.ref#5.4.2. which are the so-called exceptions said in this rule: In all cases except the last (i.e., implicitly converting the initializer expression to the referenced type), the reference is said to bind directly to the initializer expression.

It looks like we consider initializing an rvalue-ref from a prvalue a "direct reference binding", for better or worse.

In any case, I'm not seeing anything that could be fixed editorially here (i.e. if we change something, it will have effects on the meaning of the standard).