Reply #14: The difference is...

[Bill Bored]

...if the lever machine is tested, the "pencil lead" problem, and just about any other problem, will reveal itself in the process. If a piece of software is tested, it may never reveal problems that could manifest themselves during the subsequent election, or in a previous one.

This leaves the post-election audit as the only way to verify an election (chain of custody issues notwithstanding).

That said, NYVV has never been very strong on election verification, IMO. It's mostly been concerned with the usability of paper ballots vs. DREs, which is the "voter verification" argument. And I think it's a pretty weak one given how easy it is for most people to use a lever machine. (Others in NY are of course free to use electronic ballot markers to mark paper ballots and NYVV deserves a lot of credit for helping to make that happen.)

On edit: with respect to the exit poll true believers, they have their own crosses to bear, but I'm sure it wouldn't take much imagination for them to come up with a theory that the exit polls in NY were deliberately skewed to make the levers look bad in an effort to replace the last transparent voting system in the US with HAVA junk. In fact, all they'd have to do is read this post!