Reader Jerry Whitmarsh took me to task for passing along the CNW Marketing Research study on energy costs of various cars without due statements of concern over the methodology. He kindly provided me with this site. The criticisms seem to fall into three categories.

First, the energy cost imputed for production and recycling are too large: 'NW's figures, for example, show that the Civic Hybrid can cost nearly $165,000 more over its lifetime, "dust to dust," than the standard Civic, which is a difficult figure to swallow, even considering the extra development, materials, and disposal of the Hybrid variant. Honda's Integrated Motor Assist (IMA) system is a mild hybrid system and many engineers have admired its elegant and simple design and function, considering the efficiency gains.'
Second, CNW understates the expected life of hybrids: 'For instance, according to the study the average Prius is expected to go 109,000 miles over its lifetime, while a Hummer H1 would go 379,000 miles. CNW says about hybrids: "…these are generally secondary vehicles in a household OR they are driven in restricted or short range environments such as college campuses or retirement neighborhoods." '
Third, CNW does not distribute the research and development costs over enough units: 'One other area of the study that some critics have found to be misleading is that CNW only included the so-called design and development cost of models sold so far, not on the potential volume of that technology in the long run.'

Similarly, an earlier study relating to CO2 emissions rather than energy use reports that hybrids score better on average than gasoline-engine cars.