The guys who did the analysis are not climatologists, and David says that the journal they published in is well known for favoring papers that are skeptical of global warming. Still, the authors say their data set and methodology are public and they “welcome” scrutiny.
I don’t know how important this graph is to the overall climate debate these days, and in any case the authors may have found some genuine errors or they may turn out to be entirely full of hot air (so to speak). David has some additional background about this on his site and promises to follow up as responses come in from the climate community, so check back at his place periodically if you’re interested in the latest updates.
UPDATE: Probably a false alarm. More here.