I did suggest that you shouldn't use Rivara and Thompson. There was a reason why I advised that you not reference them. Using the same data as Rivara and Thompson you will also find that helmets reduce skint knees.
Check the para on Problems with the Seattle Study
They conflate all kinds of injury. I accept that minor scrapes can be reduced but serious injury is increased.
You're assuming there is no down side to wearing a helmet. This is not supported. Evidence shows that the incidence of serious injury increases where helmet wearing has been made mandatory.
Opinion is not very strong evidence. I used to use a helmet, I read more about it and found it was a poor choice so I moved on.
Are you disputing the drop in mileage in Australia? Some of the information on that comes from road side surveys and some comes from the national census. They show the same pattern from different sources. This is a good thing to have in evidence.
Australian figures being 20 years old does not invalidate them. Helmets were actually made to a higher standard then.
Here are some more recent things.
Magazine article about helmets and decline in standards over the last 20 years.
And consider this. Every time you see a face cream ad on TV there is a wee line about how 7 out of 10 women in a sample of 80 through it had some effect. I drive Volvos and for years their USP was safety. They would put some data in their about collisions at 30 mph, SIPS, crumple zones and so on. There is big money to be made selling helmets. If there was any good evidence in favour of them we would see it but we don't. I had a look at Specialized's web page. Their helmets are known to be the best for standards compliance. I didn't find any claims about their effectiveness outwith compliance with standards. They talk about weight, ventilation and comfort but they do not try to sell their magic hat safety device on evidence of safety. Why would that be?