In this post, he relates changes in the VIX to changes in the S&P (a % change in the VIX results in about a 10 basis point change in the S&P in the opposite direction), with an R-squared of about 50% (i.e. variation in the VIX explain about 50% of the variation of the S&P). He also notes the mean-reversion in the VIX, and that the S&P tends to be high when the VIX is low (and vice-versa).And yes, he's been added to the blogroll.
He lays out some of the math of VIX mean reversion here (caution - serious nerd alert ahead), and gives a few applications here (including a slick way of using implied vcolatility to get beta).
Some Interesting Uses For The VIX
David Merkel at The Aleph Blog has put up several great pieces on the VIX (the CBOE Volatility Index):