Yes, certainly. My point was more that the upper bound of what the law allows tends to be less than what the most skilled and experienced road users could safely do, because most road users aren't that skilled and experienced.

For example, while I am certainly expected to slow down on a motorway if there's driving rain and limited visibility, I am not permitted by law to exceed 70mph under any normal circumstances. This remains true despite the fact that I happen to have many years of driving experience, to have received more training than most drivers, and to drive a more capable car than most drivers, all of which correlate negatively with accident risk and might suggest that I could drive safely at faster speeds under appropriate conditions.

You're trying to rebut an article neither I nor the article was making. An "Idaho Stop" permits a cyclists to treat a stop sign as a yield and a red light as a stop. If a pedestrian or automobile is already in the intersection or will get there before the cyclist, the cyclist must yield, including coming to a full stop if necessary.

I appreciate your point, I just think it's rather like saying a speed limit isn't a target. While it's true and the spirit in which the rules are intended, the reality is that most drivers will assume driving the limit is reasonable unless and until it very obviously is not. Moreover, driving significantly below the limit and holding up other road users without a good reason is widely considered antisocial, but again you have the problem of not everyone having the same judgement about when slowing down is sensible.

There are plenty of other problems with driver judgement, too. A lot of drivers maintain that they can be talking safely on a mobile phone while driving, perhaps because they have a hands-free system installed, despite overwhelming evidence to the contrary. But in this case, our law explicitly allows for using a hands-free system while driving even while prohibiting a hand-held device, despite the evidence showing very little difference in safety (both versions being horrendously dangerous). That law has been openly used by both arrogant drivers and the marketing of hands-free system manufacturers to "justify" the dangerous behaviour.

So my concern with changing the rules for cyclists at stop lines and red lights is not that it wouldn't actually remain illegal to do dumb things, it's that more cyclists might be encouraged to do those dumb things because as they see it they appear to have the law's blessing and it sways their judgement, and that the harm caused by this might outweigh any good caused by allowing responsible and safe cyclists to make smoother and faster progress at times.

Cars, bicycles, and pedestrians all have completely different behaviors and properties. I think laws should be designed to maximize safety and usability of roads for each class of users (while balancing all those interests, which its the tricky part).

We also agree on all of this. Sometimes different rules are appropriate. There are lots of streets where I live that are one-way for motor traffic but have a small, contra-flow lane for cyclists allowing them to go the other way as well, for example.

I just don't necessarily accept, on such evidence as I've seen so far, that where we do have multiple modes using the same road space in an inherently dangerous area like a junction, it's the right time to have different rules for different traffic.

And this a good example: given there is concrete evidence that Idaho stops make things safer for bicyclists, it doesn't really matter what you or I "feel" is true about the issue.

And if that's the evidence in Idaho, then good for them, I wish them every success in making their roads safer and their laws better. But I also think we should be careful about extrapolating too far. A lot of awful road designs and traffic laws get made because someone had an idea, perhaps even a good idea under the right circumstances, that was taken too far and applied too widely.

Cities with large volumes of cyclist traffic tend to have less accidents per capita because drivers and pedestrians are used to them being around. Laws that make cycling more safe and enjoyable tend to make people ride their bikes more. They also signal that municipalities that adapt them view bicyclists as more than just second-class users of roads.

Again, I agree with your general points, but again I suspect there is also more to it. For example, as well as comparing accident rates in places with more or fewer cycles on the road, it's also interesting to compare rates in places where some specific allowances have been made for on-road cycling against places with different kinds of dedicated cycling areas. We've had only limited success trying to separate cycle traffic from heavy motor vehicles here in the UK, because often it's done half-heartedly for political and/or financial reasons and just winds up creating an "alternative" for cyclists that is still inferior to cycling on the main road. However, some of the more comprehensive schemes in places like the Netherlands do seem to get good results, in some cases by explicitly not putting cycles in with motor traffic at junctions but also by recognising them in their own right and providing things like dedicated phases at signal-controlled junctions. Even if the Idaho stop idea proves to be safer than traditional everyone-stops arrangements under some conditions, that doesn't mean we should stop considering other alternatives that might be better still.