Since you've already tried removing all negative feedback, I'd start there again. Then lift one leg of both R38 & R39, thus disconnecting the feedback to V1 from V2. I can't quote chapter and verse from RDH4, but for those that don't know, that feedback to the cathode of V1 achieves similar results as a bypassed cathode would - namely increased gain and lower output impedance. BTW, Earle, these are also from the same spots you were probing.
If this doesn't help, I would then add a ~300pF, 1kV cap from each of the EL84 plates to ground. Though since you've already stated you've disconnected all your feedback, I doubt this will help in your case.
Finally, do you have all the grounding attached directly to the PCB? An external star ground could be causing some strange return issues. Disconnect any external chassis grounds in your tripler circuit and have them directly attached to the PCB.
Shannon