Yeah, it wasn't at all clear to me from 2 minutes of static analysis under what circumstances an email would be sent and to which address. I'd hope the authors of the linked article would have verified this with a packet sniffer (as mentioned in another comment here) before making the claim.