It's like if car crash injury/fatality statistics failed to distinguish between licensed & trained drivers and those who have no / suspended licenses, etc. (It wouldn't hurt if education on how to bike safely in cities were more common in the US, either.)
Also, living somewhere with a lot of cyclists (such as Amsterdam) means that people are more likely to learn how to cycle safely from other cyclists, and that drivers are more likely to understand how to share the road with them. There's a major network effect.
That's per year, your lifetime risk is ~50x that. Or, .0005% * 50 = 1 in 4,000 which is a fairly large reduction in risk of death. Granted being a heavy smoker is ~1 in 3 chance of early death and plenty of people still smoke but this is one of those things that can kill you when your young.
PS: I would also assume that as the numbers of hours / year increases so do your risks. I don't think the 80 million people average more than ~2-5 hours per week riding so more riding time = more risk.
Also, that's 1 in 4000 before 60 (60 - 10 = 50). You have a 12% risk of death before 60 which means if you died before 60 the there is a 1 in 500 shot that it was biking without a helmet that killed you. Granted a significantly increased change that it was biking, but using a helmet would not have saved you.
I would agree that biking after 60 without a helmet does not really increase your death chances that much.
According to that site, having everyone wear a helmet could prevent 250 to 500 deaths each year out of 80 million bicycle riders (700 deaths, two-thirds with brain injury, 45 to 88 percent prevention with helmets, 80 million riders). So if you wear a helmet, there is a half of a thousandth of a percent chance that you will benefit from it.
When you look at it THAT way, the safety improvement sounds trivial, nowhere near the size of the benefit from wearing a seatbelt. It gives a whole different perspective than "45 to 88 percent of brain injury can be prevented by a helmet".
And, that site pulls data from "multiple sources", not all of which may be legitimate, and appears to be a pro-helmet propaganda site.