LWG-3543 Definition of when `counted_iterator`s refer to the same sequence isn't quite right

[StephanTLavavej]

LWG-3543 Definition of when counted_iterators refer to the same sequence isn't quite right

In #1965, A. Jiang (@frederick-vs-ja) categorized this as:

Implemented at first, only comments should be updated.
_Same_sequence performs as least increments as possible, which is still OK for current wording

The comment is:

STL/stl/inc/iterator

Lines 1278 to 1284 in 6fe2eae

	template <common_with<_Iter> _Other>
	friend constexpr void _Same_sequence(
	const counted_iterator& _Left, const counted_iterator<_Other>& _Right) noexcept {
	// Per N4861 [counted.iterator]/2, two counted_iterators x and y refer to elements of the same sequence iff
	// next(x.base(), x.count()) and next(y.base(), y.count()) "refer to the same element." Iterator equality is a
	// fair proxy for this condition.
	if constexpr (forward_iterator<_Iter> && forward_iterator<_Other>) {

I agree that the comment should be updated to cite WG21-N4892's new wording. (It appears that this isn't a case where the comment should be removed entirely.)

[AlexGuteniev]

I don't like the new wording, as according to it, the current _Same_sequence may be claimed wrong, as it does not try each possible n that might exist, instead assuming a specific n.

Otherwise, looks good-first-issuish.