It's a common belief nowadays that on-base percentage trumps batting average and slugging percentage as the best predictor for scoring runs. However, in recent years, slugging has made a comeback, and is currently more important than getting on base in terms of scoring. (Batting average is still last among the three main stats: BA/OBP/SLG.*)
I looked at ML-wide batting stats from the 2009 season through yesterday. Hitting for power, not getting on base (or hitting for average) correlated most with scoring runs while batting average correlated the least. For example, the Yankees are only 10th in batting average this year, but third in OBP and SLG, and come out second in scoring. In 2009, Colorado ranked 16th in batting average but sixth in OBP and SLG, and correspondingly ranked sixth in runs scored. And last year, we know about how Toronto was a great power hitting team (first in SLG), but they didn't hit for average (ranking 23rd), but finished a respectable ninth in runs scored.
This may be a new trend in baseball that began with PED testing in 2006. Runs per game have decreased every season since testing began (what a coincidence!), as has SLG, and they're both currently at their lowest level since 1992. As power hitters become tougher to find, the ones that do exist become more valuable. I believe we'll see this trend reflected in personnel moves over the coming years. From evaluating prospects and amateurs to free agents and trade targets, GMs will put a premium on power hitters like they used to on "OBP-machines." (Ahem... Brian, a guy named Jesus says hi.)
* Bear in mind, this was only meant to analyze the triple slash stats (BA/OBP/SLG) in regards to scoring. OPS, which combines OBP and SLG, is still better than any one of the triple slash stats.
If you're interested, here's the Excel file.
This was written a while back, but Frank beat me to the punch. Serves me right for procrastinating.