Regarding Ellsbury and the leadoff spot, Glanville has an op-ed in today's New York Times discussing leadoff hitting in general and Ellsbury and Red Sox in particular, in case anyone is interested.
nytimes.com/2009/09/21/opinion/21glanville-leadoff.html
Glanville argues that there is more to hitting first than OBP. Specifically, he suggests that Ellsbury's speed and what not induces errors, fielder's choices and poorer pitching that aren't directly credited to Ellsbury's stats.
Are there any attempts trying to measure this threat effect? For example, how much, if any, does the likelihood of a homerun increase when Ellsbury is on first base compared to when he isn't? Or compared to say Drew?