This reinforces how the rampant spread that was going on during the first vaccine trials helped to speed up the process. It was so much easier to have lots of infected people in the control group when there were 150,000+ daily infections than now with 13,000.
If we only had the original strain today, with the current vaccinated adult population, and the reduced community spread, that benefit vs risk calculation would be different. The big question is, how does the delta variant change that calculation?
We could assume delta is enough worse to tip the scale. We could assume community spread isn't going to stay as low as numbers suggest in the specialized populations that kids congregate, schools. We could wait for actual numbers to show if either of those assumption is correct or not. Areas at the high end of vaccinated population will have different math than areas at the low end. They could approve it with nuance around community spread in areas the kid will be. They could approve it on the assumption all those things are true and want a uniform recommendation. They could wait for the numbers to roll in, if they're not bad then no foul. But, if they're bad, that's real people that could have had different outcomes.
It's definitely not a simple decision and the current environment it's being made in doesn't look the same as when the original EUA for over 65 was done.
If this was a cancer drug and every one of those kids was at deaths door, the math would be way different. Reinforces how careful vaccine approval is vs just about every other thing.