My one real worry is the use of the term "reducing uncertainty". The palaeo-world has become a much more complex place in the last 10 years and with all the
different calibration methods, data processing methods, proxy interpretations - any method that incorporates all forms of uncertainty and error will undoubtedly result in
reconstructions with wider error bars than we currently have. These many be more honest, but may not be too helpful for model comparison attribution studies. We need to be careful with the wording I think.