I know an email address that appears in this leak but not in any other leak that hibp knows about. That might invalidate that hypothesis (or alternatively hibp's dataset might not be as complete as it could be).
I wasn't implying this was just LinkedIn data, apologies if my tl;dr of the article wasn't clear.
You should just read the blog post.