BTW it is not difficult to design a splitter circuit for transformerless AC mains input. What is required for the simplest case for a 1:3 divider is 3 big caps (connected in series) 3 power mosfets in parallel to each cap in the source follower configuration (2 n-channel and 1 p-channel) and 3 resistors to provide low current reference divider for mosfet gates (plus bridge rectifier). This gives you a circuit that provides 3 buses of 110V*1.4/3=52V. As long as these buses are loaded equally there is no power dissipation in the mosfets ! (explanation below) In reality there will be some variation in loading but it will be small compared to the charger power that is assuming number of cells /active converters is a multiple of 3. If it is not - keep reading (some math required) -
A bit of explanation (for those less invovled in EE):
1. Assuming 3-way bus splitter let us call bus currents I1, I2, I3. There are 4 wires total feeding the converter buses connected to 3 series capacitors. According to the Kirchoff's current law the currents on these wires will be I1, I2-I1, I3-I2, -I3 respectively. If the currents I1, I2 and I3 are approximately equal (lets cal it I)the wire currents will be I, 0, 0, I which means the middle points of the capacitor chain gets no DC current. All current comes from the mains. This condition means that input (splitter) balancing power dissipation is zero.
2. Each DC converter presents itself approximately as a constant power load (at least on a short time scale) Vi*Ii = Pconst. That is at a given output power the lower is the input voltage the *higher* is the input current. This means that balance condition (1) is inherently unstable - the higher is the voltage difference in the power spliter the larger is the current forcing the splitter out of balance. This is called unstable equilibrium (like a trying to balance a ball on a top of a larger sphere) - therefore an active servo is required to maintain equilibrium. Yet as long as the equilibrium is maintained there is no power dissipation by the servo circuit.
3. If power loading on 3 buses is not equal but nearly constant - let's call it P1, P2, P3 then adding a condition Vi1 + Vi2 + Vi3 = Vtotal = 110V * 1.4 = 155V, then requiring Ii1 = Ii2 = Ii3 a system of equations can be solved yeilding exact voltage split ratio needed to satisfy equilibrium condition and therefore no power loss in servo circuit ! This means it is perfectly OK to have 3 converters on bus 1, 3 on bus 2 and 4 on bus 3 for a 10s charger. If a moderately higher voltage is supplied to bus 3 then the equilibrium condition can be attained. This simply means that 3-resistor voltage divider must be approprietly calculated.
4. If power loading on 3 buses varies dynamically (what really matters is the power ratios which do not change much in our application) then a dynamic servo can be designed using current sensor on two intermediate buses with goal to nullify all intermediate bus currents by adjusting mosfet current and resepective bus voltage. As long as servo succeeds in doing that there will be no power dissipation on mosfets.
I've thought of some circuits for the dynamic mode but I do not really think dynamic tracking is needed. A bit of balancing current through mosfets for static case is not critical.
. 
