Did you not read my post? "One might come up with ways to surveil the booth (whether in general or through the coerced voter), but at least we have a chance at detecting this."
We have a chance at detecting purely electronic hackers also. For example, hackers made 574 attempts to connect to one of my servers as root since the logfiles were rotated.
This idea that physical is somehow categorically better than electronic is just magical thinking.
Let's say I'm an abusive husband. If my wife has the choice of voting online, I can force her to choose online voting, make her vote at home and in front of me, and nobody has any way of detecting my coercion. If my wife has no choice but to go to a polling booth, election observers absolutely do have a chance of detecting my coercion.
> This idea that physical is somehow categorically better than electronic is just magical thinking.
No, it's a demonstrable fact. You have created an "electronic hacker" strawman here. The problem I am raising is that of coercion, not a man in the middle. You have not been able to provide any means of mitigating it when not using a physical polling booth.
Problems such as "electronic hackers" are only problems on top of the problem of vote coercion, which is clearly made much worse with any ballot system that does not use physical polling booths.
You won't go very far by tampering with a single vote. Try coercing with 10 thousand people, and see how easily you are tracked.
The benefit of online votes is that coercion and data stealing are the only flaws we must take care of. Instead of this huge structure trying to cover for all the flaws of paper, we can focus on those two well specified ones.
Again, let me point out communist terrorists have already hacked this. Selfie in the voting booth. "Chance of detection" is just an assumption that some magic occurs because things are physical.
Your hack also works for absentee voting, which we already have. Do you propose eliminating that as well?
The good thing about physical is that it demands much more effort to tamper with results.
Also, the mechanisms we put in place to prevent tampering are easily understood by just about anyone, not just people with CS degrees - which lends credibility to the process, which I find to be a benefit.
