In a recent, wide-ranging and lengthy interview (which includes an embedded 36-minute interview video), Richard Nisbett (Mich--Psychology) explains his deep skepticism of multiple regression. Nisbett, a leading psychologist, makes his position crystal clear. For example: "A huge range of science projects are done with multiple regression analysis. The results are often somewhere between meaningless and quite damaging. ..." Also: "I hope that in the future, if I’m successful in communicating with people about this, that there’ll be a kind of upfront warning in New York Times articles: These data are based on multiple regression analysis. This would be a sign that you probably shouldn’t read the article because you’re quite likely to get non-information or misinformation."
The full interview implies that much of what bothers Nisbett, on a technical level anyway, involves omitted variable bias. To be sure, an important problem, but itself does not wholly discredit regression analysis altogether. Even if it misses the mark in (important) places and risks over-claiming, Nisbett's critique (or, in his words, "crusade") nonetheless warrants attention.