Completely agree, the ST70 sans the input pentode is also a highly cloned circuit, typically a 12AX7 up front in common cathode stage coupled to a split load phase inverter to drive push pull finals. Sadly since they ditched the input pentode they lose a lot of gain that was once turned into feedback so the modern triode input stage clones have worse distortion figures. I get it, pentodes can be noisy so why not use a dual triode cascoded to behave like a pentode? That way there you get a similar open loop gain figure to keep feedback about where it was in the original circuits.
That's my latest issue, I was asked to build a SE amplifier for highly efficient horn speakers and although I was using a 6AU6A as the voltage amp stage there was just too much chance to get a noisy 6AU6A. Even screening for good samples I reduced the noise floor by ditching the pentode voltage amp stage for a dual triode cascoded. My opinion is with low power amps like this SE KT88 build, I am getting a clean 12 watts with distortion just under 1% THD, I am using a 5670 with both triodes in parallel as a cathode follower DC coupled to the KT88 to get higher peak power in A2 operation. The 5670 with gm doubled has a fairly low output impedance at the cathode and seems to be doing an ok job of driving peaks into grid current. A mosfet would work better but he doesn't want SS in his amp and doesn't want to fork out money for interstage transformers. The KT88 is run as a pentode and not UL or triode strapped. I am working out the current feedback from speaker for an adjustable damping control that allows for both +/- current feedback to give different output characteristics. Even though the amp can output power higher than 12 watts that's all I rate it for because it's not really meant to be run in grid current mode for long periods of time, with the efficiency of his horn speakers (105db at 1 watt) he won't be using more than a watt or two average but there is more headroom for dynamic range compared to just A1 operation.