My comment is definitely an oversimplification that takes into account the rest of the context of the article which outlines all the differences which makes the author's work transformative enough to count. You're right that derivative works aren't a straight up protected use as my comment stated. However, "Not at all" is wrong, because derivative works are protected within the doctrine of Fair Use provided the work satisfies enough of the four-rule doctrine to qualify [1]. Transformative works are a subset of derivative works which are protected [2].
Just to clarify my grandparent comment here, I was thinking "transformative work" when I said it was protected by Fair Use, not "derivative work" as I stated. So I was wrong!
To clarify my sibling response here, I'm acknowledging that statement was wrong. However, to say that _derivative_ works are "not at all" protected by Fair Use isn't entirely accurate either, as they absolutely can be depending on the the judgement and application of the four-point doctrine of Fair Use. The parent comment here cites part of the copyright law, but Fair Use, which I cited, is a core component of the same copyright law that limits the application of the section they cited.
"Only the owner of copyright in a work has the right to prepare, or to authorize someone else to create, an adaptation of that work."
https://www.copyright.gov/circs/circ14.pdf