I do statistical reporting for a living, and first off, kudos to everyone who's posted for critical thinking! Stats are given way too much weight, with way too little thought given to where the data came from, how they were generated, what definitions were used, or who's funding the study.
Second, the original article was utter crap.
Third, the actual study looks pretty straight-forward to me (thanks for posting it, dsjr70, you rock!). The lack of a conclusion isn't a bid for more funding -- at least no more so than any other scholarly study -- but rather an acknowledgement that they only looked at certain factors, and so can't draw conclusions based on data they didn't look at. For example, they can't say ER folks are racist, 'cause they didn't measure their attitudes towards race.
To me, the implication is pretty clear. The researchers couldn't responsibly say it, but I can, since I'm now officially just some guy on the internet. It looks like an income/geographic location (i.e., hospital) thing to me. My guess is that african americans tend to make less money than whites (even after controlling for insurance status), and so live in poorer neighborhoods, and so when they crash they tend to get taken to worse hospitals. I'd guess that each hospital treats blacks and whites the same, but the ones with fewer resources and higher patient counts get more black patients, and the nicer hospitals with more individualized care and higher-paid, maybe even better educated doctors get more white patients, just due to the fact that rich people are are more likely to be white (or vice versa, if you prefer). I'd be very curious to see what the results look like when controls are added for income. Given their methodology, it would be tough to control for geographic location, but I'd also be curious to see how motorcycle crash death rates (or any death rates) for hospitals correlate with proportion of patients who are black. I imagine it's kind of the same thing as how inner-city schools seem to have both fewer resources and worse educational outcomes.
Sorry for the long post -- I get excited about numbers. I'll just finish up by saying that certain factors -- income, ethnicity, and health problems, for example -- all tend to have relatively strong correlations, so it can sometimes be difficult to tease out which factor is playing the most important role.