4 deaths per 50 million miles means it’s just under 1 death per 10 million miles. That would put riding a one wheel for 1 mile at the same risk as biking 1 mile, walking 2 miles, or driving 25 miles[1]. That is of course only looking at death and not serious injury.

Still, as other people have mentioned there are solutions Future Motion has been avoiding that would significantly lower the risk. Adding small wheels on each side in case of nosedives, or adding a loud audible warning in case of low battery or max speed being reached rather than just a subtle physical alert are two that have been suggested here and seem like they would make a big difference.

[1] https://en.wikipedia.org/wiki/Micromort

It looks like most death rates for transportation products are measured in deaths per 100 million passenger miles. So 8 deaths seems like a reasonable estimate.

By comparison, cars are about 1.5 deaths, and motorcycles are 25-30.

For the car number, I'm willing to bet that OneWheel riders skew younger and male-er than the general driving population. I'd bet the risk difference mostly disappears if you're comparing similar demographics.

For the motorcycle number, keep in mind that you can't legally ride a motorcycle without a license and some training and/or non-trivial testing. And most riders wear much more extensive PPE than the average OneWheel rider. And they're still at something like 300–400% of the risk.

I think it kind of comes down to whether you think of a OneWheel as a toy, a recreational product, or a form of transportation. It's ridiculously unsafe compared with, say, a boardgame, and probably on par with recreational products like skateboards and jet skis, and somewhere in the middle between driving a car and riding a motorcycle for transportation.