> The fact that there are bad bike lanes is not a reason to stop building bike lanes, it's a reason to build better bike lanes.
Perhaps, but I think it's worth recognizing that, counterintuitively, a bad bike lane can be worse than no bike lane.
> The modal shares of cycling in the Netherlands or Denmark (segregated provision) vs, say, the UK or US (largely no distinct provision) are a pretty clear indicator that segregation gets more people cycling.
Not convinced; there are many differences between these countries. I've heard it argued that e.g. much greater legal liability for drivers who injure or kill cyclists is what makes the difference.
The modal shares of cycling in the Netherlands or Denmark (segregated provision) vs, say, the UK or US (largely no distinct provision) are a pretty clear indicator that segregation gets more people cycling. If the collateral damage is that a few roadies have to slow down and use bike lanes in cities, I'm not uncomfortable with that, even speaking as a confident on-road cyclist myself.