Combining the coercive potential of vote-by-mail with internet network-mediated vote casting is still not a good idea.
Similar to the rush to purchase inadequate voting technology after HAVA was passed and funded, this is just a bad idea until we have made some significant strides in a number of elements of basic science (regardless of who funds it).
People interested in these ideas should concentrate their efforts on kiosk-based remote voting models.
(BTW, it doesn't help to "feed the trolls" and having a back-and-forth here in this forum on this issue with this person is exactly that. Save your breath (hands?).)