Have we improved our confidence in climate sensitivity since 1970?

Warren Washington co-authored the first paper I know of with the phrase "cloud feedback" in it, and it implied a very wide range of uncertainty. Then early work with the NCAR 2nd generation GCM showed regional teleconnections would complicate things further. The Charney report in 1979 offered the famous 1.5 to 4.5 C warming range that endured for nearly 30 years. Now where are we, how soon might we significantly narrow this range, and how is this information used in assessing the efficacy of various policy options to mitigate climate change?