A two-dimensional solution to a three-dimensional problem
It's possible that many municipal workers and civil engineers become entranced with the plan views on their blueprints and don't consider the real-world applications of their decisions. A case in point that I see everywhere is the first bike lane marking that appears outside an intersection. I maintain that these are placed too close to the intersections to be of any use for those turning onto the marked roadway, likely just because someone with a two dimensional overhead drawing saw that new traffic may enter a given road at an intersection and determined that the logical spot for the bike/chevrons graphic is, of course, just at the exit of the intersection.
The trouble is that, from the driver's perspective, the visuals are hopelessly distorted - made abstract - by the driver's position when negotiating the turn. If the driver is looking well ahead, where he/she should be, the bike lane marking is certainly not recognizable for what it represents (assuming it isn't missed all together.)
The proposal: Just move the first post-intersection bike lane graphics further down the road, so that drivers can find them once they have finished the relatively intense action of navigating the turn (No pedestrians, cyclists, or other vehicles hit? Check!) and are able to properly process any new information that comes from driving straight forward. Easy peasy.
This example is for a left-turning driver, but I'd posit that the same issue applies to right-turning.
I'm not saying adopting this will make cycling any safer, but it will certainly help municipalities communicate most effectively on their limited budgets. These suckers can cost over $100/apiece, so bang for the buck really matters. May as well ensure every driver sees every marking.