Of ten nanotech-for-the-environment predictions from 2005, the one with no named material — better lighting saving 200 Mt CO2/year — blew past its number by 2x. Almost every specific mechanism stayed in the lab.