Reading through the linked paper, it seems that as long as P is fixed before the proof is attempted, wFS is secure. I probably could have made it more explicit,

Reading through the linked paper, it seems that as long as P is fixed before the proof is attempted, wFS is secure. I probably could have made it more explicit, but that is indeed my assumption in the setup of the blog post.