I think this is the source data for those viewgraphs:
http://avt.inl.gov/hev.shtml

A ton of great information- lots for any camp to chew on for talking points, as bwilson said. The failure rates have been shot around quite a bit, but many of the naysayers seem to not mention a few items:

- these vehicles were integrated into existing gov't fleets, which are in general driven far harder than personal vehicles. Think 'free rental car.'

- the sample size is very small, to the point of being almost statistically invalid.

The data is fine as far as it goes; just remember to take the above points into account. What this presents is somewhat better than anecdotal evidence, but not enough to make assumptions about any of the models or technologies in many cases.