Exactly. If you want to reuse the term "entropy" in information theory, then fine. Just stop trying to make a physical analogy. It's not rigorous.