Based on an interview with Alan Hirsch

The interview with Alan Hirsch can be found on Fangraphs here.

Let's begin.

"First, there was his belief that statistics alone could identify top prospects. Jeremy Brown and Brant Colamarino are a good place to start if we’re talking about shortcomings in identifying undervalued assets through statistical analysis."

Another good place to start would be to judge Billy Beane's acquisitions of minor league and major league talent.  I am throwing out an idea out there.  I don't think the belief was that in statistics alone; some of the idea was there so that they could save money on the draft.  For instance, Jeremy Brown was a first round pick but had to agree to a lesser contract (read: signing bonus) upon signing with tthe Oakland A's.  There are some shortcomings when analyzing amateur athletes and projecting them into the Minor Leagues and eventually the Major Leagues even if you use traditional scouting methods.  In your previous response, you talk about the success of the Twins, who are usually defined by using traditional scouting methods.  However, I could easily pick out players that they have drafted that would indicate that we should already be talking about shortcomings in identifying undervalued assets.  Furthermore, it's not as if the Twins select undervalued assets.  The majority of their picks the lst few years have been "toolsy, athletic" outfielders and pitchers who throw strikes instead of walking a lot of batters.  Here's just a small sample of failed first round picks by the Twins.  Chris Parmelee, Matt Moses, Adam Johnson.  The conversation might need to be, why is it so difficult to project amateur players into professional settings?  The success rate in the Major League draft is incredibly low.  Perhaps there needs to be a revamping of how teams view the draft and how to project those same players.

"That pitchers have some control over the outcome of batted balls is established if you look at pitchers’ career data, particularly comparing pairs of pitchers who spent most of their careers with the same team (thus controlling to a large degree for ballpark and defense). Pitchers are a complicated package. They succeed as a result of different combinations of skills reflected in walks, strikeouts, home runs, AND weaker-hit balls that generate easy outs."

Well I'll look at that.  The Twins certainly believe in this philosophy that a pitcher gets in trouble if he walks players.  Part of the reason for this I believe is that the BABIP is much higher with players on base because it's easier to find a hole when a player is slightly out of position.  I haven't done enough research to prove this, however. 
John Smoltz, Greg Maddux, and Tom Glavine were teammates with the Atlanta Braves from 1993-2002.  They are probably our best examples of looking at paired teammates who played for the same team and everything.  Smoltz's BABIP during that time ranged from .268-.300, Maddux's was .244-.324, Glavine's was .248-.320.
Smoltz's low was .268 which was in 1993, that year Maddux's was .269, and that year Glavine's was .280.
Smoltz's high was .300 which was in 1998, that year Maddux's was .262, and Glavine's was .265.
Maddux's high was .324 which was in 1999, that year Glavine's was .309, and Smoltz's was .289.
Maddux's low was .244 in 1995, Glavine's was .281, and Smoltz's was .288.
Glavine's high was .320 in 1994, Smoltz's was .271, and Maddux was .253.
Glavine's low was .248 in 1997, Smoltz's was .297, and Maddux was .280.
Their career numbers for BABIP are similar Smoltz at .283, Maddux at .281, and Glavine at .280.
It doesn't seem like the pitchers have much control over their BABIP as they are very similar even when we account for the fact they were pitching with basically the same factors each year.  If pitchers are able to control weak ground balls and such shouldn't Maddux and Glavine have much lower BABIP's then Smoltz?  Glavine and Maddux both much lower K rates then Smoltz and so would need to generate more outs in the field, however, their BABIP's are only slightly lower than Smoltz's.  It seems that while they were playing with the same defenses that BABIP is largely luck and that pitchers do not have control over it.  I'll revisit this later mainly to find a better way to look at it.