Since trading began in 2010, the trend for BTC volatility appears to be downward:
http://www.coinometrics.com/bitcoin/bvix.
This is strong evidence that as adoption increases, bitcoin will become less volatile.
is the timeframe long enough, and the error on the slope of the line small enough, that this evidence is even significant?
When I ran the numbers through R, the trend was statistically significant.