Also, I differ with your "Given the few times that moving a runner up INCREASES actual scoring" comment.
Swapping an out for a base may increase the potential for A Run to be scored but, over time, decreases the potential to score multiple Runs in an Inning. Very simply, trading Outs for Bases is a net Run Value loss proposition.
Non-choice Productive Outs (meaning we're excluding bunts) occurred at a rate of only 2.1% of all Plate Appearances in 2004. Knowing that Base Hits PLUS Sac Flies were the result of only 24.3% of all PA in 2004, we can use that information to understand that a random ball hit into play Productive Out to move a Runner is truly beneficial AT MOST in 0.05% (yeah, half of one percent) of all MLB Plate Appearances. We also know that Productive Out behavior decreases the overall number of Runs scored by an Offense over time.
Now again, that's AT MOST 0.05% of all PA. Considering that not all Base Hits score runners advanced by a "Productive Out", you've got yourself a big bunch of nothing.
And knowing THAT, it's not at all surprising that three sabrmetric-driven teams (Toronto, Boston, Oakland) finished with three of the lowest four slots in all of MLB in "Productive Outs". They've figured it out.
The other team that finished in the bottom 4? The Cincinnati Reds, who finished 20th in MLB in Runs Scored even though they had fewer Base Hits than all but three MLB teams and led all of MLB in Strikeouts.
Seriously, if Strikeouts and swapping Outs for Runners were primary offensive drivers then why did the Reds finish 9-10 spots higher in the MLB Runs Scored standings than you'd think they would?
Because neither K's nor Productive Outs have pretty much nothing to do with Run Scoring over time.