From this video
We can simplify the circuit to be a single capacitor with capacitance of 50 micro farads.
Q = CV, we can determine the charge on each capacitor in series is 4800 micro-coloumbs.
So far so good.
The confusing part:
The 1st capacitor has a charge of 4800 micro coloumbs, as expected
The 2nd "capacitor" is actually a parallel circuit. We calculate the charge of C2 (70 microfarads) using (7/10 * 4800) = 3360 microfarads.
This means the charge of the other path (C3 + C4 in series) is 4800 - 3360 = 1440.
Still, so good.
However, since C3 & C4 are in series, they both must have the same charge. So they both have 1440 micro-coloumbs!
This means the total charge on the circuit composed of (C2,C3,C4) ... adds up to 3360 + (1440 * 2) = 6240, which is greater 4800.
My understanding of what is going is: C3 + C4 should be treated as 1 big capacitor, so the 1440 charge is on both of them at the same time rather than 2 instances of 1440 micro-coulombs on each individual capacitor.
Does this make sense?