`"\u{A8}".normalize("NFKC")` should return `"\u{20}\u{308}"`, but it leaves the string unaffected. It works properly with non-Latin-1 characters. I will submit a patch.
Pull request: https://github.com/WebKit/WebKit/pull/24808
Committed 275062@main (c1a059873be7): <https://commits.webkit.org/275062@main> Reviewed commits have been landed. Closing PR #24808 and removing active labels.
<rdar://problem/123328161>