I get frustrated when I hear talk about BABIP and how it will regress to 0.300. I previously posted a link (a year ago) where the author shows that BABIP is much higher for line drive hitters than ground ball hitters. Barney has been hitting line drives instead of seeing-eye ground balls (ground balls are more a matter of chance and infield alignment).
Although this doesn't relate to Barney's case, home run hitters BABIP excludes home runs, so BABIP can't be considered to be a fair correlation for home run hitters.
If one wants to correlate BABIP to batting average then BABIP should be grouped into BABIP for predominantly ground ball hitters, BABIP for line drive hitters, and adjusted BABIP for all hitters to include home runs.