Spoonamore's simplistic (and obviously flawed) theory is that a man-in-the-middle attack was used to modify the vote counts during transmission from each county's tabulator into the central tabulators. Such an attack would work only if no one at the county level knew what the local tabulations were. But the county level officials
did know what their local tabulations were. Each county Board of Elections went through a canvassing procedure required by law under which they reviewed their county's totals, made various manual adjustments to them (like counts of provisional ballots) after they had come out of the local tabulators, and then sent their official tabulations in to the Ohio Secretary of State independently of any use of the tabulator to tabulator links. These certified totals were not transmitted tabulator to tabulator. They were faxed in, mailed in, called in, etc.
I'm not much of a skeptic -- I do believe that the 2004 election was stolen in Ohio. But the theory presented by Spoonamore is ridiculously flawed. Since the local county officials read the local tabulation results directly off of the local tabulator machines, the only way to have rigged the tabulation would be to rig the local tabulators. That could have been done by direct manipulation of those machines or, perhaps, by a remote connection to them. The latter approach could perhaps have been supported by the network architecture in the diagrams that Spoonamore has alluded to, but such an attack would not be a man-in-the-middle attack and would be significantly different than Spoonamore's stated theory. Spoonamore (and Arnebeck, Fitrakis, et al) should perfect the theory because the one they are proffering is obviously flawed.
BTW, lala_rawraw included David Dill's explanation of this flaw in her excellent article early in the time line of the Spoonamore story:
"It seems that the major concern is whether routing election results through a third-party server would allow that third party to change the reported election results,” Dill wrote. “These diagrams haven't answered my basic question about that idea. The individual counties know the counts that they transmitted to the state. If those results were altered by the state or a middleman, I would think that many people in many counties would know the actual numbers and would raise an alarm."
http://www.rawstory.com/news/2008/Documents_reveal_how_Ohio_routed_2004_1031.html