Hmmmm, I am still thinking about the issue ... but my first guess is that for high density areas a "separated" bike line is a positive change. I will try to address all of the points in the prior posts, but time is limited so I may miss some and I will try my best to keep the statements/thoughts organized.

The problem the separated bike lanes address--at least in the video--is not high speed impacts. It is more to do with keeping traffic flow, minimizing low-speed impacts and avoiding injury on heavily used streets. Given the high density traffic, the real worry is not the initial injury--probablistically speaking, one could still get a nasty head injury at low velocities--but getting run over subsequent to falling. Ignoring intersections for the moment, I think that separated bike lanes address these issues well.

I agree that cyclists will have to ride slower to ride safely since pedestrians, roller-bladers, and such will either use the lanes or cross during "jay-walking" and so on. I think that this is the price of congestion, living/working in a high density area and cooperating with our "neighbors" in a civilized fashion. A civil engineer is probably trying to optimize something that includes travel time/safety of all commuters/travellers. Generally, if less accidents occur and flow is improved, then I would expect the average commuting/travel time to decrease for the average driver/cyclist. Note that the italics emphasizes that it is an empirical question. The underline emphasizes that most people on this forum are not average cyclists.

But I don't agree that this necessarily leads to an absolute loss of rights to riding in the streets with bike lanes. I rather not get into a long argument about politicial science and public choice (since these assertions are difficult to prove and easy to rant), but in brief, if more people cycle then as a concentrated lobbying group cyclists become more powerful. Moreover, the removal of rights usually requires more evidence/effort than the enabling of rights. My caveats: These are not absolutes and are more anecdotal thoughts about the world than based on rigorous research. However, I would not be surprised if cyclists were not allowed on roads with separated bikes lanes during rush hour.

I think that it is easy to write/say that enforcement of traffic laws should be increased. However, in practice, strict(er) enforcement is more costly and difficult than it appears. Remember, not only does an officer have to take the time to stop and write the citation, but there are administrative and potential judicial costs as well. Mind you, I agree that agressive driving should be more strictly enforced. But I just want to point out that as the transgressions become more minor, roughly speaking, enforcement becomes more costly from an opportunity cost perspective. I will note that it could be the case that (A) strict enforcement of bike lane laws could be less expensive and/or more effective than (B) less enforcement with separated bike lanes. However, that is also an empirical question.

Intersections ...

The post that states that one cannot block part of the right lane when turning right makes a good point. And I don't know enough about civil engineering to intelligently propose a structural and cost-effective solution to the issue. Maybe some sort of speed bump? I don't know ... ... but I also think that for the intersections where right turns are a big problem--when you have multiple lanes turning right, that the solution to take the right lane doesn't help much either.

I recall that there are two methods for making left turns. One is to slide over into the left lane and make the left. The other is to cross the intersection, turn at the following corner, and then proceed forward when the light turns green. Remember that these separated bike lanes would probably be used on the busiest of streets. Think of 5th Avenue. My guess is that a large proportion of cyclists are not sliding across with the traffic to make that left turn. So perhaps some cyclists are slowed down with the installation of separated bike lanes; but in the aggregate, I doubt that the loss is significant.

I need to get some work done, but let me end with thoughts on prior research and cycling data. In my casual review of the research and available data--mind you, I have a real job to attend--I just can't see making many strong conclusions or "hard" statements about the empirical reality of cycling. In other words, the quality and detail of the data is unable to support the strong opinions I read on the forum and elsewhere. A good "for instance" is the following: there is a literature that states that cycling on the sidewalk is more dangerous than on the road. I find that much of the research fails to adequately account for selection biases in the data. Suppose that sidewalk riders are less-skilled than road riders. Or that people use sidewalks in more dangerous areas compared to the areas where they ride roads. Using the metrics in these studies, one could easy conclude that sidewalks were the cause of the danger when the true underlying cause was something else. Please, let this not turn into a discussion about sidewalks ... there is another thread for that. I can make a similar argument about bike paths.

I don't want to "poo-poo" the body of research. There are carefully crafted and well-thought out studies. And I certainly have not done an exhaustive review of the literature. I just fail to see how strong, conclusive statements can be based on the underlying data since we are unable to model much of the relevant behavior.