This is a hypothesis that the Met Office seems to be testing in the series of papers they have released today. There are three documents:
Having focused on climate sensitivity in recent months, as far as this blog is concerned it's the third paper that is most of interest. As readers know, there is a surfeit of new observationally constrained papers that have found low climate sensitivity. Strangely, the Met Office authors only consider the Otto et al study, which had a relatively high ECS estimate - a function of the ocean temperature dataset used. As we know, if any other dataset had been used then they would have got an estimate in line with the other recent observational estimates.
Sure looks like spin to me.