I don't dispute their findings, though neither do I confirm them.
However, if you take this into account, it's a whole new ball game. What about the forms of energy - electricity, which IMO is more what they use in manufacturing, is somewhat easier to come by than hydrocarbons. That is, you can get electricity by renewable sources if you like, though mostly it is still using fossil fuels right now, whereas fossil fuel powered cars can only run on that. It's not just the cost of energy (which I believe they're using as an analogue for the amount used), per se, that's the issue, but also the form that it takes.
Their findings are probably accurate, since, as I mentionted, IIRC a large proportion of electricity comes from fossil fuels at the moment; but neither are their findings going to remain constant. Perhaps it will even change when the amount of hybrids produced increase.
Not to mention the ever-present possibility that they've overlooked something in their calculations (just try to imagine the difficulty of assessing whole-of-production energy use, from component manufacture to assembly..)