Then Worldometers is miscalculating somewhere. Or it's possible the over 90's have a 30% death rate, which would give a 20% overall death rate (I guess - I mean, I'd need to see the calculations). Since there are far more older people dying, we'd have to multiply their rate by their actual number of cases (but somewhere someone has to be dying at a way higher rate than 14.8% to get to an overall death rate of 20%).
Also, that's not proper data for the 0-9 group. See here (Table 2). The CDC doesn't report by the age brackets mentioned above so I assume WorldoMeter is doing extrapolation (which I teach my students to do, it's a valid thing to do as an information strategy).
At any rate the US has lost 9 children 14 and under as of 4/25. 4 were under the age of 1. That illustrates my earlier point about...infants are more vulnerable, those deaths under 1 year were mostly under 6 months, although I do remember a 1 year old with serious underlying conditions dying).
But 9 != zero. And worldwide, the number of children dead is greater than 9, so still not zero.
It's true that children are way less likely to die of CoVid19, so far. The Kawasaki-activation syndrome going on in UK is very very concerning). But that syndrome is in keeping with all else science is finding out about this novel Coronavirus.