The latest wrench is a new paper published early in January titled On the reliability of the U.S. Surface Temperature Record. If you are like me and have done any looking around skeptical sites, you are exposed to shot after shot of US weather stations that show poor siting factors followed by claims that the data is hopelessly corrupted. This paper has looked at the actual data involved, not just the photos of the stations.
The authors took the data from the poorly sited stations and compared it to the best sited stations and found very little difference (hat tip to John V who did this years ago). However the efforts of Mr. Watts and his team of volunteers has not been in vain. The paper does detect a small problem with the current temperature trend - the current trend is actually a little (very little) biased towards the cool. This means that the temperatures are actually increasing slightly more than thought.
If you are interested in the details, John Cook over at Skeptical Science has a nice analysis and I encourage you to visit. Now, this is obviously not the last word (there seldom is a last word in science), but what is interesting about this is that since the paper has been published, the responses should be published as well. Reading those will be interesting.