You conflate election fraud (hacking machines), voter fraud (individuals illicitly casting ballots), and state-sponsored misinformation campaigns. Those are three distinct problems with very different challenges.

You say "no evidence of $x != evidence of (not x)" which is true in theory, but let's step into reality: "No evidence, despite systems built to detect evidence".

You act as if people aren't already on the lookout for fraud, or that systems are not already resistant to them.