My question is-does it even matter? Conventional thinking says yes-"homegrown" players tend to be those who you know most about, since you've scouted them all the way up since they were drafted or signed. They're basically indentured servants for 6 years (after all of the years spent in the minors), so the money saved there is another benefit. But I wanted to see if statistics supported the assertion that homegrown players are essential to a winner. Keep in mind that the league averages hover at 35-38%.
My method-take the 13 players with the most PA's and 12 pitchers with the most IP's in the regular season from the World Series winners in the 2000's. Yes, it's an incomplete and small sample size, but it's at least a start.
2008 Philadelphia Phillies-10/25 players (40%)
2007 Boston Red Sox-8/25 players (32%) (note-a lot lower than you thought, eh?)
2006 St. Louis Cardinals-7/25 players (28%)
2005 Chicago White Sox-5/25 (20%)
2004 Boston Red Sox-3/25 (12%) (really? wow)
2003 Florida Marlins-6/25 (25%) (also shockingly low)
2002 Anaheim Angels-12/25 (48%)
2001 Arizona Diamondbacks-5/25 (20%)
2000 New York Yankees-10/25 (40%)
What does this prove? Nothing really. "Homegrown" is a little too subjective for my tastes-but there has to be a cut-off at some point. But the point is-sometimes your homegrown guys are what makes your entire team (2002 Angels). Sometimes your homegrown guys are your biggest pieces and you fill the rest of the team from outside sources (2000 Yankees-Posada, Jeter, Williams, Rivera, Pettite). Sometimes you use your homegrown guys and trade them for the right pieces (2004 Red Sox aren't a perfect example, but they did move Nomar Garciaparra and Matt Murton for Orlando Cabrera and Doug Mientkiewicz).
I think it is safe to conclude that homegrown talent does matter, just not always in a quantifiable way. Sometimes you just have to accept the fact that not everything can be quantified...but baseball simply isn't such a setting where people can accept that!