[megameter]

The most important part is to do some distillation under the principle "data is not information". When you collect a bunch of sources you have data. When you filter according to a benchmark of your choosing and bind the useful bits into notes, you have extracted information.

Getting information is hard work, and acknowledging this stops me from pursuing it much of the time. A lot of articles that proport to discuss a topic directly are not informative and try to cover up that fact. So instead you have to figure out side channels. E.g. for gleaning understanding of human meat-eating I would look for discussions of carnivores in the general sense, extract that into a set of "meat eater traits" and then search for the traits in humans one by one.