Thanks, Tango, as always, for spearheading this project.  The standard deviation scores are particularly interesting. 

It might be fun to attempt a positional weighting of the various skills (post-facto) to determine suitability for position. So, for instance, the speed/reactions scores would be weighted to be of more significance vis a vis arm scores for centerfielders as opposed to right-fielders.  This might give you some information about how Vernon Wells would rate as a right-fielder, instead of centerfield, or Aaron Hill as a third baseman instead of a second baseman. 

I've actually done exactly that in the past.

The only issue I'd have to include is putting in a huge downward adjustment for LH at 2b/ss/3b, and, I guess, a small one for catcher.