So if the manufacturer spends the extra money for a UL winding they would, seemingly, have optimised it for the tube type and circuit.
@viridian This assumes that the manufacturer knew about the patent and designed according to it rather than the tradition that developed trying to get around it. In a way, David Hafler might be the one responsible for that; he was one of the inventors when he was at Acrosound, to whom the patent was assigned. When Hafler moved to Dynaco the patent didn't go with him but he knew full well how to get around it. So its reasonable to assume that Dynaco OPTs are not optimized.
So a rational thinker, in possession of that knowledge, might conclude that the UL taps were improperly placed, causing the triode setting to work better. However, a rational thinker might also wonder if the OPT had windings to accommodate the difference in the plate load impedances required since that value for a given tube is different for a pentode as opposed to the same tube wired in triode.
I'm certainly wondering that. Whether I'm a rational thinker is another matter.