Some, but not all joking aside, the SRPP can be made horizontal by splitting the power supply and shifting the voltage relations within the circuit by moving the top triode and resistor Rak to in between the two power supplies. The bottom half remains unchanged, but the top half now goes horizontal; thus , the output, while still at the cathode of the second triode, is at ground potential. Increasing the first triode's grid voltage causes the tube to increase its current conduction, which creates a larger voltage across resistor Rak, which in turn forces the second triode's grid negatively, as the other end of the resistor is connected to its cathode, which in turn forces the second triode to decrease in its conduction. (This assumes that a an external load is found at the output to provide the alternate current path. Without this path, the two tubes must share an equal current draw, as there would be only one current path through both tubes.) The delta in current draws is delivered to the load:
     ILoad = ITube1 - ITube2.
In other words, push-pull operation. In fact, I believe the pushing and pulling action is more apparent in this topological rearrangement.
     Yes, an OTL power amplifier could be built upon this topological foundation, just as Circlotron/Atmosphere amplifier (the horizontal push-pull amplifier) was built on the totem-pole push-pull amplifier (the vertical push-pull amplifier) circuit used in the Futterman amplifiers. Twenty 6AS7s would be a nice start, as, unlike the Circlotron/Atmosphere circuit, a high idle current is mandatory to keep the amplifier always in strict Class A operation. Remember, if the first triode stops conducting, it cannot control the conduction of the second triode. Of course, the grids could be capacitor coupled  to ease the bias setting and help regain wasted power supply voltage. The load could be  coupled via nickel-core transformer.
    (Imagine how busy the patent office will be upcoming weeks, after this topology is invented by ten or twenty tube circuit designers.

    I take much pleasure in providing prior art to thwart the legal adventures of those who keep these circuits out of the public domain. And I am tempted to create a new web page titled PriorTubeCircuitArt.com, which would display old and NEW circuits that might slow down the patenting of old and obvious circuits such as the recent patent of the cathode follower directly attaching to the grid of the following tube. Unbelievable, but sadly true.)

The Horizontal
White Cathode Follower
     Once again, we split the power supply into equal voltages; and we shift voltage relationships, so that the input and output are at ground potential. This circuit resembles the SRPP in that it comprises its own phase splitter. Any deviation in the cathode-to-grid voltage causes the plate resistor to develop an in-phase signal that is relayed to the second triode, which then inverts that signal and thus strives eliminate the voltage deviation. Like the White cathode follower, the optimal value for resistor Ra is the reciprocal of the Gm of the tube used. This value ensures the greatest balanced current swings into a load. However, in those cases where a large current swing is not needed, the value of resistor can be increased to help lower the output impedance. In fact, in this case, this resistor can be replaced with a choke or a current source. And once again, yes, an OTL power amplifier could be built upon this topological foundation