Yes. This is absolutely true. People don't speed around mountain passes on snowy nights. So I'm not clear what your point is.
A better way to frame it is do we want to optimize everything for automobile travel speed and convenience? That's what the US has done for the last 100 years, and the result is poor public transit, unwalkable neighborhoods, urban cycling as extreme sport, and an epidemic of road rage.
Rather than making roads visibly more dangerous, there are better options.
One I heard long ago on a motorcycle forum is: get rid of seat belts, mount the driver in a plexiglass bubble in front of the front bumper, and put a bloody great spike in the middle of the steering wheel.
No. The implication is that driver behavior adapts to road conditions, and that slippery curvy roads demand higher attention and care to navigate. In many instances, improved driver attention compensates for the increased risks of dangerous road conditions: more "dangerous" roads may not have proportionally increased accident rates or fatalities.
In the extreme case where a road appears life-threatening, a driver with that knowledge chooses not to drive.
The most dangerous conditions occur if the road appears safe but has unforeseen or transient hazards.
A better way to frame it is do we want to optimize everything for automobile travel speed and convenience? That's what the US has done for the last 100 years, and the result is poor public transit, unwalkable neighborhoods, urban cycling as extreme sport, and an epidemic of road rage.