Changing the bias point of the 6AQ5 - in my estimation - would have no effect on the AC voltage range to the El panel, which is set by the gain reduction control. (No DC voltage is present at the El as the coupling cap (C11) blocks DC.)
So, regarding the effect of the plate (R34) and cathode resistor (R36) of the 6AQ5 in each revision, the only impact would be on the tube itself, and from what I gather looking at the voltages of the various revisions listed in Jahsen's book, the 6AQ5 current draw at idle is actually higher with 10k/470R, thereby eliminating longevity of the tube as the motive for the change. Thus, the move to 10k/470R was most probably made only to facilitate a more linear operating range for the 6AQ5 - at higher gain reduction settings.
Lastly, as the latest revisions (1968 forward) use 10k/470R, it would be reasonable to assume these values represent best practice.