By routine profile, I meant "of the data we all already know the NSA has access to, such as google search history/mail content/etc".
e.g. they may have been on an 'automated search' list for being at three-degrees-of-'terrorist'-kevin-bacon and then got on a 'physical visit' list for trips to china + a few keyword hits in the web activity logs. at which point it seems obvious they would ask about pressure cooker bombs, even though they weren't there just because someone had searched online for pressure cooker bombs.
Now, I'm certainly not defending the NSAs list-making justifications nor accuracy. The whole thing is bullshit.
But this story looks like a textbook use of the bullshit programs we (now/all) know about. And it gives us no reason to theorize the existence of some new (trivially unsustainable, useless) program.
As for if they got it from "old" or "new" NSA programs - I don't think anybody really cares at this stage. What one would care is if his next google search would lead to a 3am SWAT visit with flashbang grenades or just a friendly visit from a bunch of armed guys in SUVs. That's the worrying part, not the name of the program the government uses to make it happen.
We already know from the slides that they do searches of their ill-gotten web traffic against a laundry list of keywords. Probably, every single remotely terrorist-sounding keyword.
I'd imagine "Ruby Ridge" and "fertilizer bomb" and "jihad" are in there as well. But the agents only need to mention "pressure cooker bombs" because that's what they'd actually gotten hits on.
The question is whether it makes more sense for this visit to have been the result of google alone (unlikely-to-implausible) or was the result of being pre-targetted for other reasons, which then brought their web activity under scrutiny (which fits everything we already know about the NSA's processes and policies and the facts in this case).