Fix overall statistics computation

Resolves #236 (closed)

We used to have two separate objects for computing data quality metrics: DataQuality and DataResolution; the former computes statistics for the entire MergedData object, and DataResolution computes the quality in resolution shells, each with a limited d range. DataQuality was incapable of specifying a d range, making refreshing the overall data quality after changing d min/d max problematic. Now the "data quality" or overall statistics are computed as a DataResolution object with one resolution shell, the entirity of the specified d range.