There was a candid statement about climate models that were made at the Advanced Study Program/Early Career Scientist Assembly Workshop on Regional Climate Issues in Developing Countries held in Boulder, Colorado on 19–22 October 2011.
The Workshop is reported on in the April 3 2012 issue of EOS on page 145.
The relevant text reads [highlight added]
One recurring issue throughout the workshop was that of managing complex impact assessments with a large range of results from global and regional models; variations between models are often not fully understood, accounted for, and/or communicated. Also problematic is the discrepancy between the spatial and temporal scales on which regional climate projections are made (tens of kilometers and ~30–100 years) and the scales that are of primary interest to many communities in developing countries (kilometers and 0–10 years) that are presently affected by climate change.
My Comment: I agree with this comment except I would delete “change” in the last sentence. Climate is always changing, and the use of the word “change” itself miscommunicates the actual threats faced by developing countries even with climate they have seen in the past.