I think that drawing conclusive conclusions (try saying that ten times fast...) is difficult largely because a well-controlled experiment has never been carried out, and would probably be considered unethical (where 1/3 eats only McD's; 1/3 eats the exact same menu/macros, but with quality ingredients; 1/3 is a "control" and eats a "normal" diet), given the hypothesis that at least one of those diets is really bad for you.
Because most of these studies have been epidemiological in nature, there is a lot of correlation that may not equal causation. For example, the study suggesting an increased correlation between red meat consumption and cancer. Aside from having to rely on subjects' memories of what they ate, there's no controlling for whether they grilled their own grass-fed ground beef or got McD's, or how it was prepared (i.e. sugar-ladden bbq sauce vs. simple salt and pepper; butter vs. canola oil).
This site has been a great resource for me:
www.marksdailyapple.com
Some of what's said in the nutrition forums is a bit nutty, and nothing resembling scientific, so take with a grain of salt.
But hey, it's gotta be useful on some level, as the fitness forum where I first read about Starting Strength...