I don't think RBIs are a meaningless stat, but I don't thing it tells all that much either. I much prefer to look at their splits with RISP to determine the quality of hitter in RBI situations. I agree that the #1 goal on offense is to not make outs, but in RBI situations, I think that driving that run(s) in becomes more important (productive outs).
JD Drew is an interesting hitter. He isn't someone who consistently performers well with RISP, he seems to flip flop every year, having a great year followed by a terrible year. 2009 he was terrible with RISP, 2008 he was great, 2007 he was awful again, 2006 great, 2005 awful, 2004 great, and so on.
Now, there is a correlation (really causation) between guys who perform in the "clutch" and how many runs they drive in. When looking at 2009, it would appear that JD Drew had a fantastic year and that his low RBI total was low due his place in the lineup (few RISP opportunities). While it's true he had fewer opportunities than Pujols (143PA to 189), Drew squandered his opportunities where as Pujols made the most of his.
Now, a lot of people don't believe in "clutch", but I think it's simply a misunderstood idea. I believe in clutch, but I believe in it over a large sample size, not per single at bat. There are guys who do well in pressure situations and guys and don't. The guys who consistently perform well in pressure situations are what I would call "clutch".
I love having JD Drew in RF, but he is too much of a Jekyll & Hyde to consistently drive in 100+ runs if he were to hit in the 3/4 hole. He's not the prototypical run producer.