This doesn’t actually analyze HIBP in any way. In this case, HIBP is just an inspiration for this article.

This algorithm just checks a password you enter against the Levenshtein distance of words in the password database, and if you are close enough, it tells you that there is a match.

That’s two totally different things.

And the HIBP database is huge compared to anything the author might have dealt with here.

Moreover, algorithmic complexity of calculating Levenshtein distances on huge databases has not been considered.