This applies to some journalistic endeavors as well. In some cases, I seek out the thoughts of those whose facts are unreliable, because their opinions are interesting or novel. But just as often, I seek out long-form, descriptive pieces to learn what reliable people believe is true. I cannot validate every aspect of the outside world, so I trust others to do this. When they are frequently right, like Nate Silver, say, I trust them more. When they lend their brand to more bizarre groups, I trust them less, not because the sponsored link itself changes the truth of the rest of the articles, but because it signals a lack of judgment that might have a common source with many other important and difficult decisions a magazine or newspaper must make.
For instance, when I read the Economist, I feel rather confident in their facts. They have a particular fiscally conservative slant, but in general they have proven to be realistic and relevant. They might not espouse particularly novel solutions, but they lay a solid framework for further thought. As it is, The Atlantic is reasonably trusted. That's almost surely the reason Scientology would like to place a sponsored story there, in addition to, or in lieu of, more popular outlets.
In the coming transformation of journalism, institutions like The Atlantic, or newer upstarts like Svbtle, will have to consider sponsored posts and similar "brand-lending." When they do, they'll have to decide whether they want to be a trusted brand, or an interesting one.