Forgive me for saying it like this, because it's always difficult when one's whole line of academic work is being called into question, but isn't Chara right about this? Am I not right about this? (writing from this other account that I have because this is an important issue hence I'll take the risk of being accused of mere block-evasion.)
Isn't the best solution to managing the risk of using software to deal with elections simply...not using software? What's actually wrong with voting in a physical way? There's plain *advantage* to physical voting: physical evidence of voting. What's actually gained by using software? Speed.
Should elections be decided quickly? Is that actually a necessary value to assert?