Right - would you rather rely on a theorem which the author claimed is true, but won't let you see the proof, or one where the proof is published and reviewed? That's the choice between closed and open source math software.