That's a good point, Peter.
Here's a MINT 9 copy of that variation with the "missing pixel" clearly shown and, as you said, it's non-existent on the Goodwin card...
And here's an example of a card with the marks on the holder being magnified by the Goodwin scan (this is the scan from the Goodwin page and I remember confirming with Bill that the white marks were not on the actual card)
