Add description of differences between a single declaration with multiple bindings and multiple declarations with single bindings

[benblank]

The Semantics section of the README emphasizes that using x = expr1, y = expr2; (multiple bindings in a single declaration) is only approximately equivalent to using x = expr1; using y = expr2; (single bindings in multiple declarations, but the same bindings in the same order) — and similarly for await using — but I didn't see a description of the practical differences between them. I also didn't see an existing issue addressing the differences and during a quick scan of the specification diff, I didn't notice a clear explanation, either.

To me, the emphasis on "approximate" equivalence implies that, while the differences between single-declaration/multiple-binding and multiple-declaration/single-binding semantics are not commonly relevant, there do exist differences which are situationally relevant. If that is the case, I feel that at least a quick summary of the differences and their potential relevance is warranted. If not, I feel that implication introduces confusion and that the emphases should be removed.