This is really splitting hairs, but i think it's:

48% of bombing victims/people who died from blast injury are children

of those children 40% were under 10 years of age

so .48 * .4 = 0.192 meaning roughly 20% of bomb deaths were under 10.

But like if you're having this conversation you've already lost. There's no way to frame it so it's not horrific.

The text does not say "of those children", the text says that 48% of the whole are children, and 40% are under 10 years. I agree it's a little ambiguous, but I read that as meaning that 40% of the total bombing victims were under 10 years.

It says the dataset is 'colleague's households' which might be different from all of gaza.

The "among our colleagues' households" is the key part. It's not generalizable to the whole of Gaza.

I'd assume that "victims" includes injured, not just killed.