Not that this comes naturally, but...
Huhne does have at least some points....and that's from someone who's not only seriously pro-nuclear, but took his bachelors degree in Nuclear Engineering.
We have managed to consistently cock-up technology choices - and then make an even worse job of actually building stations in the past.
Going with the Magnox designs was probably sensible, given the constraints of the time - little enrichment capacity, no forging capability for larg(ish) high pressure vessels and the fact we'd got the hang of machining high-purity graphite when we built the Windscale piles.
What WAS unforgivable was the fact that in building eleven stations, we used seven different designs. Which completely bolloxed the idea of a learning curve.
It was only then that we REALLY cocked -up. In the early 1960s, the CEGB and SSEB approached the then Secretary of State for Energy for permission to start the procurement of their "second generation" nuclear plant. They thought the gas-graphite line of development had gone as far as it reasonably could - and wanted to adopt an American LWR design. The UKAEA was at the same time lobbying for a "super-Magnox" - the AGR (funny that an R&D organisation wanted the option that needed lots of R&D, but maybe that's just me being cynical). The CEGB claimed the complexity of the design, and the need to fabricate almost the whole thing on site would make it expensive and prone to delays. The Secretary of State sided with the AEA, as he wanted to develop a "distinctively British" technology for export - that was Tony Benn, by the way, fresh from the decision to build Concord.
Not a single AGR was ever sold abroad. Not a single AGR was finished on less than a 50% cost over-run, or less than three years late on a seven year build programme. Hardly surprising, because not only had we picked a pig of a design to build, we then managed to decide to build three and a half different designs amongst seven stations built. Dungeness B, the first to be started, was twelve years late on a seven year build. THe last two, Heysham II and Torness were meant to be built as copies of Hinkley Point B and Hunterson B, but finished up with change on every major system.
At roughly the same time as we opted for the AGR, the French were faced with the same choice. They went the other way, and licensed the Westinghouse PWR. In the same time it took us to built seven AGR stations (14 reactors, totalling 9,200 MW) the French deployed about 55 PWRs, totalling about 50,000MW, using just four designs, each larger than the last. They also manged to export something like 20 of the things.
We're about to build a derivative of that design, scaled up to 1600MW in a single reactor - the EPR. The other candidate for new build in the UK is also a PWR derivative - the AP1000 from Westinghouse .
Only one country has built a gas-graphite reactor in the period since we started building AGRs.
The moral of the story? Keep Ministers, and especially the British Civil Service, a thousand miles away from any aspect of technology choice.
Next, onto Thorium