There is no NEED for E-voting. 12-24 hours to handcount paper ballots is sufficient and also enough to have the counting audited/supervised by independent parties.
The problem then becomes, 'How do we determine who is an independent party who is unbiased enough to give us a truthful audit?' Other than that little problem, though, I agree with you fully.
However, according to discussion in the Foundation of Mathematics e-mail list, archives of which are available here, the members of the prize committee were "informed but not polled" as to the validity of the proof. The prize committee members were Lenore Blum, Greg Chaitin, Martin Davis, Ron Graham, Yuri Matiyasevich, Marvin Minsky, Dana Scott and Stephen Wolfram. On October 26, Martin Davis wrote to the FOM list that "The determination that Smith's proof is correct seems to have been made entirely by the Wolfram organization. My understanding is that the I/O involves complex encodings."
On October 29th, Stanford computer scientist Vaughan Pratt wrote to the Foundations of Mathematics list that the universality proof of the (2,3) Turing machine was flawed, asking "How did an argument containing such an elementary fallacy get through the filter?" Pratt points out that the fallacy of the proof could be used to "prove" the erroneous statement that a linear bounded automaton is universal. The text of Pratt's email is available here."