Earlier quoted context omitted.
Nah. The ISP is the actual medium whose network was used to copy the alleged infringing content. So that's not analogous to all the utilities in use by the infringer.
I think the key differentiator is that Cox [somehow] knew that the customer was using their network to copy the alleged infringing content. The electric company doesn't know this, and (to use another example) the hard drive manufacturer doesn't know that their customer is storing the content on the drives the make. If I were Cox, I suppose I would be incentivized to operate kind of like a VPN and just not store logs…
There's also business and legal reasons why ISPs need to log and inspect traffic. "No logging" means "no abuse desk" - i.e. if someone's using your ISP to send out spam or hack people, and you aren't logging or inspecting traffic, then you can't investigate that spam or that hacking. At a minimum it means every other online service is going to distrust your users - i.e send them through reCAPTCHA hell; up to an outright ban. Think like what it's like to use clearweb services through a Tor exit node, and now imagine that for a bunch of nontechnical users. They're going to switch to the ISP that doesn't result in them getting group punishment. But actual criminal actors are going to flock to the service, because they need someone in the middle to launder their packets for them, which makes them have a worse reputation, and so on.