I think the point is that it's not unreasonable for any sort of red-flagging system (whether data-mining or someone's gut) to have a false positive in this instance, because odds are pretty good that's what this is. If you're the FBI, the confluence of Uranium purchasing, in-depth info gathering on nuclear facilities and watch-words like "hacker"[1] and "nuke" probably should raise eyebrows. The reality is that this is the sort of situation that we'd expect to trigger a false positive even in the most competent and well-run agency.

False positives are frustrating and time-wasting, but the reality is when you're trying to find something as infrequent and stealthy as terror cells, you're going to have a lot more false positives than actual hits. We have to trade off false positives against the risk of false negatives (not detecting actual terrorists), and so long as competent police-work confines the impact of the false positives to these sorts of quick and semi-intrusive visits it's not that bad. Obviously if the FBI screws up and ends up with an intrusive and wasteful investigation here that's a problem.

[1] - I'm aware we're trying to reclaim the word, but it still has a certain evil connotation in the public mind.