Comment by dcode

That's not the issue / impression. The issue is that well-formed UTF-16 is rarely used in practice. All of JavaScript, Java, C#, Dart, Kotlin etc. effectively use WTF-16 for compatibility and performance reasons, and that's what's semantically distinct from UTF-8. These have asymmetric value spaces, so that strings in "legacy" languages would sometimes throw or implicitly mutate synchronously. Mixed systems typically use WTF-8 as the common denominator for this reason, i.e. not UTF-8, but Wasm decided against.