I think the more likely issue is just a matter of bias, not deliberate malfeasance. It's really hard to see errors in something when you get the results you expect. For TOBS adjustments for USHCN in particular, a mathematical artifact from the way they try to correct for it seemed more likely. That's based solely on my intuition as a mathematician though. Since we know they have engaged in deliberate efforts to conceal disconfirming evidence, and to punish journals for publishing papers whose conclusions they didn't like, I don't consider malfeasance completely implausible.
It's interesting to just plot the trendline of various sets. I especially find it interesting when you do so from the year I suggested previously, 2002.
CRU Global Average is how they described in on their page. Going by the link, it was HADCRUT3. Which shoes only
I meant to include this in my previous line. The way NOAA wrote that note in 2008 did not exclude peak to trough periods of 15 years or longer. It's a reasonable exclusion to make though, so I decline to hold that against them.