I fiddled around with the ratio of $SKEW to $VIX, and it looks like it is somewhat predictive. Using an adaptive moving average (default Kaufman parameters) it appears that the crossovers of the ratio with the moving average do a decent job of keeping one out of major market drops. I seriously doubt it is any better than any number of other indicators, and is likely prone to whipsaw and delay, too. Just food for thought. (And no, I don't really have a terribly rational basis for looking at the ratio. It's just the kind of thing I like to play with.)
None of the charts posted go back very far. I think this becomes a lot more interesting when you look at both 3m and 1m skew and go back farther, let's say to 2009. Check out the charts below using data from www.getvolatility.com . The August rout pushed 1m skew to the previous 2010 and 2011 peaks, but right now we're pretty much near median levels in both. I agree with longthewings, as the absolute level of vol shifts higher across the board, it's harder to maintain high levels of skew. Thought this long lookback in the charts was very useful.
Nice, charts. If your one of the devs, can you send me a pm to join the beta? I have provided an email already.