I guess I did not express my question clearly so I will try again.
The chart for a Q of 12 and my input Z says I need a two gang cap of 150 pF each.
I next use this capacitance to calculate the inductance required to resonate at the freq of choice.
My question is:
What cap value do I use? Total 300, half 150 or half of that (75) as the series value? Then, what inductance do I get? The whole coil end to end or half from CT to one end?
You calculate the number of turns on the coil it takes to resonate end-to-end with 75 pf maximum capacitance, for a balanced tank circuit. The reason you need less capacitance is that each tube is working into only half the tank circuit. Think of a single-ended final like the BC-610. The final tube's plate output feeds one end of the resonant coil, but the circuit is mid-tapped via the frame of the two-section 150/150 pf plate tuning condenser. The opposite end of the tank is essentially floating free, supplying the out-of-phase rf voltage for neutralisation, but there is no rf voltage source working into that end. So the entire tank circuit is like a 1:2 turns ratio step-up transformer. Its loaded resonant impedance end-to-end is 4 times the impedance the tube plate is looking into, because the tube is effectively tapped down on the coil, using a capacitive voltage divider.
So for the Q of 12, you need only 1/4 the capacitance you would need with an unbalanced tank circuit where the tube is working into the whole tank circuit, with the bottom side simply grounded at rf.
You can use the same capacitor for a balanced or unbalanced arrangement. If you modified a BC-610 to run with grid neutralisation, or perhaps changed the 250TH to some tube that did not need neutralising, you could wire the two 150 pf sections of the plate tuning condenser in parallel, which would give a total of 300 pf. But you would have to reduce the number of turns on the tank coil to maintain resonance.
The end-to-end rf voltage across the balanced tank coil is twice what it is across the unbalanced tank coil. Remember, the "dead" side of the balanced tank is used to generate rf voltage equal the plate output voltage, but 180º out of phase with it. That adds up to twice the voltage that appears from ground to the plate of the tube.
You could just as easily use a 75 pf variable, with double the plate spacing, but the problem is maintaining balance. The coil could be grounded at the mid-tap, but the split stator capacitor is a much more practical means of assuring perfect balance at all settings, particularly if several bands are to be covered.
If the amplifier uses two tubes in push-pull, nothing changes. If the amplifier is running class-B or class-C, only one tube at a time is ever supplying power while the opposite tube is "just sitting there" in the cut-off mode, and the tubes take turns each one supplying power to its half of the resonant tank circuit.
I rarely bother to calculate the required inductance of the coil and then use the formulae to calculate the number of turns. It is much easier to simply determine the required tank capacitance using the charts in the Handbook, and then use trial-and-error to come up with a coil that resonates at the desired capacitance setting.
But it is very important to do this with the capacitor standing alone, not connected to the rest of the circuit. I have found that inevitably, once the dimensions of the coil have been determined using the outboard capacitor, that when that same coil and capacitor are wired into the rest of the transmitter and the tubes are inserted into their sockets, the setting for the resonant point is anywhere from 1/2 to 2/3 the capacitance required when the variable condenser is standing alone. This is due to inter-electrode capacitances of the tube(s), neutralisation capactances(s), and stray capacitance of the interconnecting wiring.
