Hi Kold. :o
I am one of those philistines still using the 1961-90 averages as I believe that they represent a good benchmark for the pre-recent warming climate, whereas I think the 1971-2000 are a bit of a half-way house between the last 10 years' climate and that before the warming set-in. Anyway, that debate can run and run.
Therefore, by the 61-90 averages, July 2007 was colder than March 2006. However, I accept your point on December 2001, so just 3rd coldest since Jan 97 then. The one thing on that, though, is that the cold of that month was focussed on the CET area (the METO areal series were not as cold) whereas with this month I expect the CET value to be fully reflected in the more extensive series.