Comment by cowuser666

This treats arguments like opaque tokens where we know in advance which are good and bad, and need to make sure we have lots of good and few bad. But actually arguments have content and to know whether they are good or bad, we have to engage with them.