Well, how about "enhanced" SRPP, with anode choke in the "upper" Rk position? "Ordinary" SRPP has good features (simple, direct coupling between triodes, every tube has B+/2 and the same Ia, "upper" tube is load for the lower tube and cathode follower at the same time, etc.), but Rl for the lower tube is not CCS, actually pretty low with lower mu tubes: Rl=rp+(mu+1)Rk.
If we use CCS or anode choke (choke impedance Z=2Pi*f*L)for "upper" Rk, then this high Rk is further "multiplied" with upper tube mu, and we have very large Rl (A~mu, low distortion) and very low output resistance. We can use anode output on the lower tube, too.
Schematic shows values for the "Lundahl" LL1668 choke (100H,25mA and 680 Ohms DC resistance, Rw). We don`t have to use the same Rk like Rw, or the same Uak, or even the same tubes. It`s desirable that Rw is low, with values on the schematics we`d get about 11mA through SRPP. Nice thing is that choke parameters (L,Cw...) are not that critical in this (SRPP) position like in the anode load use.
But, I don`t have any plate choke to try it in the real world. We need a volonteer to actually try it. (If we use the same double triode for lower and upper tube, then our heater supply must be "referenced" to about 90V - better use one 5687 for both lower triodes and another for upper, the upper Uh referenced to about 200V.).

• http://www.tubecad.com/articles_2002/SRPP_Deconstructed/index.html

I just posted it as a technical reference.....Colin