It appears that CRU and NCDC discarded their raw data files at some point, as described here (thanks to Hugh for link). We downloaded our data from NCDC in early 2006. Is our data "adjusted" or otherwise "improved" or "value-added"? If so, in what way? Perhaps all they did was eliminated duplicate stations and correct some of the station names.
UPDATE: The answer appears to be yes. According to the Russian Institute of Economic Analysis, only 25% of the Russian stations were used, and these were the ones in cities. Over 40% of Russian territory was not included, and this 40% is the territory that showed no significant warming in the late twentieth century.