This is precisely the problem. My suspicion is that you should be flagging individual articles, not sources. That's no better than my grandfather ignoring any hard news reporting that comes from the New York Times because it has a "liberal bias".

In the case of Twitter's internal documents, the reason you saw nothing of "real value" is precisely because Tech Crunch drew a line between releasing information that could be damaging to individuals and the company and things they could report that were relatively benign. They received much more information than they ended up publishing. Tech Crunch could have made more in ad dollars had they just published everything. They painstakingly satisfied an obligation to share information with the public and protect companies and individuals who they saw as having done nothing wrong. Whatever side they ended up on, they certainly tried to be transparent about it.

As for the opinion part of reviewing startups and their software, I don't see any other way to go about it. Most small startup companies that are covered in Tech Crunch are based on a few facts (company name, founders, employees, location, backers) and then a review of their product or service. This is inherently an opinion. There just aren't that many "facts" about startups at the beginning. There are beliefs and perceptions. You can disagree with the opinion or analysis, but I don't see how you could avoid having one.