Steve Connor: A flawed verdict wilfully blind to key considerations

A cardinal rule of science is not to cherry-pick your data to suit your hypothesis. Reading the Royal Society's long-awaited report on the nuclear fuel cycle – delayed because of the Fukushima crisis – it is difficult not to conclude that it has employed a cherry picker of industrial dimensions.

There is only a cursory mention (in one sentence) of the technical and economic disaster of the Sellafield Mox plant, which has cost the UK taxpayer £1.3bn to date, with a further bill of some £800m in decommissioning costs. If one of your main recommendations is the building of a second Mox plant at Sellafield, one would have thought some analysis of what went wrong the first time would have been relevant.

Equally, there is no mention of the delays and cost overruns of the American Mox plant at Savannah River in South Carolina. Yet the Royal Society picks the nice ripe cherry of the Melox plant at Marcoule in France, which it says could teach us a lesson in how to build a Mox plant that works. Tell that to the Americans.

And then there is the strange omission of the "third way" of dealing with Britain's plutonium stockpile – using the existing Sellafield Mox plant to make Mox fuel for disposal, which the Royal Society chose not to mention even though it must have known about it from one of its star witnesses, Professor Frank von Hippel of Princeton University. Perhaps that was a sour cherry too far?