Do past abuses of personal data point toward future abuses of NSA metadata collection?

Stephen Walt spotlights why we worry about metadata collection:

The real risk to our democracy is what this situation does to potential dissenters, whistle-blowers, investigative journalists, and anyone else who thinks that some aspect of government policy might be boneheaded, unethical, or maybe even illegal. If you are one of those people -- even on just a single issue -- and you decide to go public with your concerns, there's a possibility that someone who doesn't like what you are doing will decide to see what they can find out about you. It doesn't have to be the attorney general either; it might just be some anonymous midlevel bureaucrat or overly zealous defense contractor. Or maybe it will be someone who wants to suck up to their superiors by taking down a critic or who wants to have their own 15 minutes of fame. It really doesn't matter: Unless you've lived an absolutely pristine online and cellular life, you might wake up to discover that some regrettable moment from your past is suddenly being plastered all over the blogosphere or discussed in the New York Times. 

 In a way, though, Walt's evidence that we should worry cuts two ways:

Why would the president say no?

In a ritual that's got to be 50 years old by now, journalists bemoan and ridicule poorly targeted pitches sent by PR people. You're a Bloomberg "healthcare" reporter covering, say, large hospital chains.  You get press releases urging you to inform your readers about a promising new skin cream. What cretin would send you such drek?

Good PR people do constantly hone, refine and update their media contact lists. But they winnow down only to a point.  Those lists are initially generated by media database screening software: you punch in, say, print publications with circulations over 200,000 and reporters covering "healthcare" and get a list of, say, 700 names. You might shrink that list to half its size, but probably not to a point of perfection, because a) the "negative" work of shrinking a list is time-consuming, and b) the cost (to your potential work result) of leaving out someone who might respond is higher than the cost of leaving in people who won't.  The costs to credibility of sending poorly targeted email are incremental and cumulative (unless you're featured in James Fallows "The Glamorous Life of a Journalist" series); the cost of striking someone off of a list who might have responded to your pitch is potentially high.