Interesting and insightful methodology, I would say, having read through the details of what they do. One issue I would have with this would be that, due to its extreme volatility, base effects will be much more significant. That implies that at any given point a particular YoY number would be a lot less meaningful. However, it's a good data point and the fact that more of this type of information becomes available (e.g. the MIT BPP, now known as PriceStats) is very encouraging.