The SRPP circuit can effectively increase the transconductance of the top tube by using a large valued Rak resistor. This increase in effective Gm lowers the output impedance of the regulator.

we make resistor Rak in value, it is still being shunted by the tube's own rp, which (when multiplied by the tube's Gm) yields the gain.
   The Cascode circuit easily can yield a gain larger than the mu of the triode. The Cascode works to shield the bottom triode from the top tube's plate resistor, and thus it works to preserve its Gm. Below is a schematic of a cascoded SRPP voltage regulator.

   Any perturbations at the output of the regulator are relayed to the bottom tube's grid via the .33 µF capacitor. The bottom tube then amplifies and inverts the relayed perturbation, which being fed to the top triode's grid, aggressively bucks the perturbation. And as we are no longer concerned with symmetrical current swings by top and bottom tubes, we can make resistor Rak   (R1 + R2) very large, which will increase the amplification of the correcting signal for the top triode. And by using dissimilar tubes, the output impedance of the regulator can be decreased even further. For example, a 2A3 or 6AS7 could be used as a top tube and a 6DJ8 or even a 12 AX7 could be used as a bottom tube.
   Still, we run into the usual limit to gain from a Grounded Cathode amplifier: the amplification factor sets the gain limit. No matter how large

    By cascoding, all of the bottom tube's transconductance can be brought to bare against resistor Rak (R1 + R2), greatly increasing the gain over the straight SRPP circuit. Consequently, the output impedance of the regulator will be greatly reduced over straight version. The DC feedback is mitigated by the voltage divider formed by the 1M and 660k resistors. Still this could make an excellent series voltage regulator.