These stats, as you say, may be "interesting" but, for the most part, are useless. Kifer's original stats are from 1978 and his most recent are 1992. That's 20-35 year old information. And you may note a pretty drastic change in the profile of the type of cyclists killed occurs during the 14 years between '78 and '92.

His conclusions are based on conjectures founded on scanty old info. To use that as a basis for how safe it is to ride a bike in 2012 probably has as much value as reading Tarot cards.