The case for "luck" with CMW still remains a strong one. With each passing year, his GB% has faded at the expense of a higher LD%. His FB% remains fairly constant. (see graph) Now, it would stupid to expect these trends to continue, as, eventually, he'd be giving up close to 60% LDs and about 20% each of the other two variety. Not likely. I would expect that somewhere close to what we are seeing this year will become the baseline.
http://www.fangraphs.com/graphs/2074_P_season_full_9_20080502.png
Line drives are usually an indicator of more quality contact against the pitcher. What's odd about this year is that, in a small one month sample, his BABIP has surprisingly dropped, which is the opposite of what you would expect. (see graph)
http://www.fangraphs.com/graphs/2074_P_season_full_7_20080502.png
That is unlikely to hold. Notice how it increased as the LD% went up and GB% dropped in previous years. For it to remain at this year's current level will require some good fortune, or, to use a synonimous term, luck.
The biggest surprise to me has been his HR/FB ratio. In his MLB years, his HR/FB rate has been 11.2%, 8.2%, 6.4%, and 3.1% from 2005-2008. This is anomalous to the trends in his contact type rates. Batters are squaring the ball up more often, yet a RH pitcher who throws half his games in the greatest HR park for LH hitters has managed to reduce the incidences where a fly ball leaves the park. Color me skeptical as to his ability to maintain this type of good fortune. His best active comparables in terms of repertoire and GB inducing ability are Lowe and Webb. Both are better at missing bats, and both have not seen an erosion of their GB%, yet both are typically north of 10% HR/FB while pitching a significant amount of games in extreme pitcher's parks (I realize Webb's home park is not a pitcher's park but the rest of the NL West is).