"there is no guarantee that your preprint is accurate with the final, published version, so most people will still trust the paywalled version more than your PD

"there is no guarantee that your preprint is accurate with the final, published version, so most people will still trust the paywalled version more than your PDF" I think this is backwards. The definitive version that should be cited is the freely-available one, since that is the version that everyone can read. No one should cite the paywalled version.