With respect Lecky - the estimate of fatalities per racing hour is a reasonable low side guesstimate of order. So "made up" is not a proper characterisation. That is the rough order of racing hours per annum, unless you can come up with a better estimate which would take a long time to do, by going back to a statistically significant group of races and actually adding up times for them. And I doubt it would differ substantially not by an order either way.
But the takeaway is that ours whilst it is a "risk sport" in need of runner responsibility and appropriate disclaimers, is nothing like as dangerous as many others, such as horse riding.
A pity that comparison was not reported by the media vultures.
I fail to understand why people regard this as any way worrying, enough to try to challenge the figures. Triathlon figures are of similar order indeed probably a lot worse, 1 fatality per order 100000 hours / 100000 outings primarily related to swimming and starting as reported elsewhere. So we are hardly remarkable.
I recollect reading that a university study had shown that cycling to work reduces average mortality by 40% - the net benefit of exercise way exceeding the risk of it. That 40% clearly refers to a subset of working age.
PS to put these figures into context, an entire lifetime is order 600000-700000 hours including sleep, now we all live longer.. Living itself is fatal eventually