Bottom line, I wouldn't feel comfortable with Andrus being rated PR unless the Texas 3b coach told me he agreed. He's an excellent base stealer, doesn't ground into dp's, takes the extra base at a greater rate than league average. This doesn't add up to a PR baserunner imo.
The problem with this statement is this: you appear to expect baserunning (and other people expect fielding) to be treated differently than batting and pitching. DMB has produced a single-season disk with batting and pitching event tables and running, stealing and fielding ratings based strictly on each player's *performance* in that single season, not on his underlying ability.

If a batter his .220 with 10 home runs, he hit .220 with 10 home runs. His performance may be adjusted for his home park, etc., but nobody expects his third base coach to be queried about whether a lot of those batting outs were the result of great defensive plays or just easy pop-ups. Nor is his performance adjusted because he had a nagging hand injury all year that prevented him from hitting his usual .300 with 25 home runs. His batting event table is based on his actual performance. DMB uses the same criteria for its single season running, stealing and defensive ratings: players are rated on what they did in this particular season, not on what they would be expected to do year in and year out.

In 2009, by the criteria that DMB uses to assess this skill, Andrus was a poor baserunner. In 2010 he may well, by the same criteria, be an excellent baserunner . . . just like a guy who hits .220 in 2009 might turn around and hit .300 in 2010.

I'm personally more interested in modeling a player's underlying ability rather than what he did in just this single season, but that includes batting and pitching as well as running, stealing and fielding. If I had any interest in playing with present-day players I'd be using the projection disk, which takes a broader look at player performance than just a single season.