BTW, since it's possible to have a qualified-id that names a bit-field, should a similar note be added to [expr.prim.id.qual]?

struct A {
    int x : 1;
    void f();
};
namespace N {
    auto [b] = A{};
}
void A::f() {
    A::x; // bit-field
    N::b; // bit-field
}

Regarding the original issue: I'd like to keep the cross-reference as-is to highlight that structured bindings could also refer to a bit-field. If you want to add a cross-reference to class.bit, do so on the first mention of bit-field in the sentence (a little earlier).

Regarding your second point: Agreed, but make that a second pull request, because it actually adds normative wording, which might cause more resistance.

1. I think structured binding should be explicitly mentioned earlier in this sentence, and it feels redundant if the same cross reference appears twice in this sentence, i.e.

The expression is an lvalue if the entity is a function, variable, structured binding\iref{dcl.struct.bind}, data member, or template parameter object and a prvalue otherwise\iref{basic.lval}; it is a bit-field if the identifier designates a bit-field\iref{dcl.struct.bind}.

(I'll send another PR for this change.)

2. I'm not sure where to put the rule on bit-field. Maybe instead of adding it to [expr.prim.id.qual], it's better to move the rule to [expr.prim.id], to avoid repetition?