I'm a little confused - 'significantly'? How significant do you think their safety contribution should become before they can be effective?
I'm pretty sure they're not likely to save me from a serious auger. There's a moderate chance they *might* prevent or lessen injuries in minor spill scenarios, and TBH that's significant enough for me because all my spills over the years have been of the minor, relatively slow speed type.
If only use things that are truly effective then a lot of people wouldn't bother with many types of cancer treatment, but they do because some chance is better than none at all, and when you're in a game of chance you play every card in the deck if you're trying to win.
If people don't want to wear helmets, then that's cool. I'm anti compulsion and can see no need for a government to meddle in our lives to that degree. But don't try and justify with arguments about effectiveness, because if the odds are 1 in a thousand that's still better than none at all.
Lets be honest - the majority of non helmet wearers, the millions of regular Joe cyclists who ride from a to b daily without a 2nd thought, who never visit a cycling forum, don't wear lids because the can't be arsed or don't want the expense. He'll, half of them can't even be bothered with lights.
So it's cool. Just be honest and say 'I Simply don't want to' rather than making general statements about effectiveness, because even the tiniest glimmer of chance is better than none.
Significantly, as in statistically detectable. I don't really care whether a helmet would save me from a cut or scrape, or other non-serious injury. I care about whether it has a detectable effect at reducing serious head injury. The evidence shows that they do not significantly effect your chance at sustaining a head injury.