The ability to swing the current between the two anodes of the 6BU8 may have been an incidental or secondary feature.
Zenith’s starting point was two heptodes, one for sync separation and the other for agc, that used the same gating signal (low-level demodulated video) on their respective grids 1. They had different signals on their respective grids 3, though. The first step was noting that heptodes were not essential, but that dual-control pentodes would work. In fact this was already established, bearing in mind that the starting point was a noise-gated sync separator circuit using the 6BN6, which was an ersatz form of the dual-control pentode. Also, in looking for lower cost alternatives, Zenith had considered the 6SA6 dual-control pentode, but the 6BE6 was lower cost still. What was needed was to combine both dual-control pentodes in a single envelope in a B9A base. Given that the same signal was used on both grids 1, then these could share a pinout. A shared cathode pinout was easy enough, as was a shared grid 2 (screen) pinout. That got the pinout count down to 9. Thus the 6BU8 was conceptually a twin, dual-control pentode, but the common pinouts made it look like a split-anode or twin-anode pentode. With the EF816, one could infer that in terms of the Pro-Electron designation system, it was seen as being basically a pentode (EF) and not a dual pentode (EFF).
Nonetheless, Zenith and GE must have been aware of the current swinging possibility, given that GE had developed the 6AR8 dual anode sheet beam tube in 1954 primarily for use as a colour TV subcarrier demodulator. (The RCA 7360 followed in 1959, and appears to have been aimed primarily at the SSB market; it was suitable for use up to 100 MHz).
And curiously, Zenith proposed using the 6AR8A as the demodulator in its early FM stereo decoder circuit:
I am not sure that the 6AR8(A) (or its 6JH8 successor) was ever used this way in a commercial product, though. GE noted that this valve was subject to stray magnetic fields, so pickup of 50/60/100/120 Hz hum could have been a problem in audio applications.
Be that as it may, clearly the 6BU8 et al and the 6AR8/7360 were interchangeable for some applications.
The 6FG7 was mentioned in the original post. This though appears to have been a VHF TV triode pentode frequency changer of the third kind.
Cheers,