There's obviously a lot to both positions, but as a succinct answer: I would argue that a mixed policy likely generates the best outcomes -- by instigating a crisis to motivate public interest and movement on the topic, and then sustained, strategic leaks over the following decade(s), once there's an interested cohort (with a baseline of evidence).
Ill date myself a bit here by saying that Snowden is the only major leak of my lifetime, but it seems to me that he was able to get traction on the issue when a lot of people had failed to (eg, there were rumors of NSA programs before including from NSA insiders, but little public attention), but what has kept the topic alive is a series of strategic leaks of new information to sustain public attention and keep the debate focused on it being a current issue.
So in some senses, I think we need those doc drops of "They're really looking at all your dick pics, hook-up texts, and tracking every illicit meet-up", but I also think that we get a lot more usage out of quieter leaks of key details, once the conversation is happening.
That was sort of the point of my analogy: quiet rumors from key staff to shareholders were ignored or suppressed by executive staff until one guy in accounting dropped the whole set of books. This obviously leaked a bunch of business secrets, but Id argue the only way to comprehend the scale of wrongdoing was to see that the entire system had become organized around that illicit activity.
It was only in that context we could even start a conversation about the scope of what had gone awry.