So, Zedza, the right way for MediaDefender to respond to the possibility that Revision3 was turning a blind eye to piracy was to destroy their site and not give them a call?

When you notice that someone has parked illegally on your street, do you complain to them, or do you just set their car on fire to teach them a lesson?

How does this method do anything except maybe identify at what theater the piracy was accomplished?

That's exactly what it does, and almost certainly the entire point. Like with most other piracy, a big chunk of movie piracy comes from insiders. If you want to get a good recording from a movie in a theater, you'll want a quiet undisturbed screening room and access to the sound. This usually requires you to know (or be) the projectionist.

So, if Paramount can identify which theater a pirated copy came from, they have a slightly higher chance of figuring out which projectionist made the copy.

It's kind of like the irritating receipt checks at the exits of stores -- they're not there because the store thinks you stole something, but because they're worried cashiers are stealing with the help of their friends. And just like with the receipt checks, Paramount has decided that the best way to fight internal piracy is by making the experience worse for everyone else.

It's a crappy way to do things, but cheaper than actually focusing your efforts on employees. And it only has a downside if you actually care about your customers and want them to enjoy their time.

I agree that it's good for an ISP to notify users if they're coming up against an arbitrary usage cap like this, but I really don't buy the "they can't do it over email, so they have to hijack your web sessions" excuse.

How does my ISP know that any individual HTTP request is something I'm going to view in my web browser? Half the applications I use daily sling HTTP requests around for various purposes that I never see -- version checks, RSS feed updates, Twitter updates, etc, etc.

Do I really want my ISP potentially screwing up one of those applications in a hard-to-reproduce and hard-to-debug way because they guessed wrong and assumed that any HTTP fetch was me sitting in front of my web browser surfing around?

It's a lousy solution from a technical standpoint, even if it didn't feel overly-intrusive and creepy.