The article lays out a whole lot of "if/then" criteria for this to be viable, and none of it seems to be particularly easy to sift through in order to connect every porn user to their habits. It seems to me like this is a lot less a matter of, "here's a list of everyone's names," and a lot more a situation where a single person could be identified and then tracked. That in itself has some scary implications that can go well beyond porn viewing habits, but I'm not seeing this as some Ashley Madison hack or Wikileaks situation, where it's a massive dump of info that's easy to decipher.
And not for nothing, but "browser footprints" aren't hard and fast. The unique configurations of browsers don't stay consistent over time, even short periods of time. And how many people stick to a single device, profile, or even physical location in their everyday surfing? The internet is mobile now. That has to play a part in the reliability of this sort of information.
I don't know. It seems to me that this might be more of a mental exercise in what could happen if a lot of variables come together correctly, and not so much a clear and present danger.