We never really exposed general population of under 34s to unmitigated coronavirus (perhaps March/April 2020 in USA or Sweden in 2020 is closest we got to it), but even that is uncontrolled for people deciding to reduce exposure to other people on their own.

"10000 less likely" was derived from 1 in 6 million compared to a global death rate of roughly 2%, then allowing for roughly an order of magnitude error: not scientific at all, I agree.

Taking your numbers as undisputable while not acknowledging that they are as uncertain as "the other side" is a bit dishonest in my opinion.