I don't think that Skye misrepresented it all that much, at least not from a reading of just the abstract.

Be that as it may, there are a couple of other bits that I see as useful. In these incidents, only 15.3% of the total crashes involved motor vehicles, and crashes with motor vehicles were 3.6 times as likely to lead to serious injury. So assuming that 15.3% held true ubiquitously, and since the NHTSA traffic stats only include motor vehicle related incidents, we could multiply the previous risk calculation by 6.5 to include the total number of bicycle accidents (ER and Hospital related).

Those who have stopped riding near traffic due to fear of accidents might feel vindicated by the 3.6 times likelihood of serious injury involving motor vehicles, although it's still a very small risk in my judgment.