How much does velocity matter?

I wanted to see the correlation between velocity and effectiveness in a pitcher. This has huge ramifications for scouting, as a pitcher's velocity is one of the first things looked at by scouts. So here I used any pitcher who threw at least 20 innings in 2011 in the NL West. I used Fangraphs customizable reports, and used FIP - as the standard of effectiveness, and used their velocity charts. My assumption before hand was that top velocity matters very little and speed differential matters much more, so I also subtracted the highest velocity from the lowest velocity for each pitcher (click on the images to enlarge).

Interesting to note that the Phillies, the best pitching team in baseball, had only one qualifying player throw 94 MPH, and no one threw harder. All 5 pitchers that threw over 95 were above average (above average is having a FIP - below 100, below average is anything above 100), so it does seem there is a positive correlation between big time velocity and success. 22 pitchers didn't throw pitches at 90MPH or more, and 9 were above average and 13 were below average.
There were only two pitchers who had differentials more than 20 MPH, and they both were above average. 20 pitchers had differentials of 15 or more MPH, 15 were above average and 5 were below average. However, 16 players had differentials of less than 10 MPH and 12 of those pitchers were above average. So perhaps there isn't much to speed differential. At the very least, we have to say the results are inconclusive. It doesn't even seem to help to have a slow pitch necessarily, as 3 of the 4 pitchers with pitches less than 70 MPH were below average.
So it does seem that having a 95+ MPH pitch correlates to success, while not being able to throw 90 correlates to failure.