The energy-use argument strikes me as ... interesting. It's not that stopping and starting takes any less energy for a car (particularly avehicle without regenerative braking), but that the rider's effort is greater. In fairness, the article states this, but the distinction could be clearer.

It does highlight just how much of an efficiency loss frequent stop/start traffic is. In my experience, my most efficient travel (long distance at ~90 kph) is about 1.7x more fuel efficient than when I'm driving surface streets with frequent stops, based on liters/100km (US MGP ratings obscure this relationship).

As a cyclist, in light traffic I tend to the Idaho Stop practice. There are a number of rationales:

• The injury a bicycle can inflict is generally vastly less than that of an automobile.

• Cyclists almost always have much better visibility and awareness of surrounding traffic, with the possible exception of traffic to the rear.

• Bicycles generally accelerate much more slowly from a stop than motor vehicles.

However, in traffic scenarios, I tend to fairly studiously adhere to the rules of the road. Why? Predictability, and my own risk. If I'm following RoTR, drivers have a much better sense of where I'll be, and I am less likely to get into an unpredictable (and likely harmful) situation.

One other item: stop signs outside the US are often substituted for with "give way" signs (you'll occasionally see "yield" in the US). Where it isn't necessary to always stop, cautioning drivers (and establishing legal rights of way) is considered sufficient. Replacing intersections with roundabouts also addresses this through traffic flow.