Maybe the point of someone else you've spoken to. Personally I don't even see how free will enters into this discussion, and if you were to ask me, I don't even think we have it as it is commonly defined/understood. But any comment on free will doesn't change the fact that Libet measured metacognition latency and nothing else.