The C4S is working correctly. (details below.)
The 431 on the A side is NOT working correctly. If you are getting 2.1v at kreg(A) then the 431 is saturated, which cannot happen if it is working and connected correctly. Normally it would be around 4.0 volts, but this value depends on the particular tube and is not precise. The experiment with the 1K resistor confirms that the tube works, it is just not getting the right bias from the (A side) 431.
Rather than speculate on what might be wrong, or get into too many detailed measurements at the 431 pines and associated resistors, I'd suggest that a new 431 be installed. Make sure the right pins go to the right holes, and that the board is clean so there's no chance of leakage currents through fingerprints or flux residue (a long shot but easy to eliminate!).
Current numbers:
The R1 value of 215 ohms gives 4.0mA, not 4.3mA as I wrote on the schematic. I apologize to everyone for failing to make those consistent. The original intent was to get 4.3mA which would have been a 200 ohm resistor, but I do not think the difference is of any significance.
The 300K resistor which parallels the tube (off the board) takes 0.6mA , leaving 3.4mA for the tube, which is what was measured. So the (B side) C4S feeding the driver plate is operating correctly.
