Can someone cite a study on this claim? I ride my bike every day and never ever see the need to match traffic, unless it happens to be going slower than me. If it's going faster ... I pick another route. I just don't think biking is safe in the city at say >17mph.
Going slow (and carefully selecting routes that allow you to do so -- the avenues in manhattan are a deathtrap btw) is the single most important factor in bike safety IMO.
I'm pretty sure the difference is not the speed but the traffic. It sounds like the odds of crashing in the first place are a lot higher in NYC than in the Netherlands.
One time I went wiped out in a bike lane because some trash bag had leaked some kind of garbage sludge all over the sidewalk (garbage day in NYC means 5-foot-tall piles of stinking trash bags on the sidewalk). The street was slightly banked, so when I hit the slick stuff my wheel just slid right out from under me and I had no time to even bail off the bike. My helmet probably saved my life that day.
At that speed, if you hit a pothole or get doored, or get hit by a car, you are going to go down very hard. I've crashed exactly 3 times on my bike in New York City, and every time I would've cracked my head open if I didn't have a helmet.