I've seen the same arguement used for sub-30mph speed limits. Talk to the anti-car crowd. They'll use that same crash severity arguement except they will use the speeds they want instead, something 30mph or lower. They generally will use the survivability of a pedestrian hit by a motor vehicle to pick their number.
Although, for my other transportation activity that I enjoy, sub 30mph speed limits would allow me to take a full lane practically any time I wanted.
Now, we can use crash severity and what people feel is fast enough. Or we can use actual data to find a speed limit value.
I find it amusing that those engineers in road building here on usenet keep insisting that their theories, their buttumptions, and calculations trump the data.
Everything I've done, I make my buttumptions, use theory, calculation etc, and when the data doesn't match, the data is what governs. I go back and see where the flaw was in my buttumptions etc. I don't deny that my design is breaking in service-testing when I am holding the broken parts in my hand. If testing shows my design holds 100lbs instead of the designed for 75lbs, I realize one my buttumptions was conservative.
I don't see why this doesn't apply to roads. A road is built, buttumptions are made and 55mph speed limit sign is put up. Years of speed data show that the average speed is 70mph and the crash rate is no higher than expected. Well, the buttumptions were conservative and the speed limit should be re-evaluated. But that's not what happens. People pound their fists insisting they know best, in a parental sort of dominince or the data is often rigged to show what is desired. (as per the NMSL era, states to the federal government).
