For a broader sample base, you just widen your category a tiny little bit to Edward Snowden, Jeremy Hammond, Aaron Swartz, Bradley Manning and Jacob Appelbaum.
Aaron Swartz was a whistleblower?
That sample set would still be statistically insignificant.
I think the point is not that one should use larger samples, but rather than the conclusion is wrong since the sample is too small. Iow, there's no proof that the situation really changed in the US on the discussed topic.
You must mean, ones who are/were made public right?