I have to disagree a little. (assume no global feedback). Viewed near the operating point with minimal signal swing, there isn't much difference.But as the signal swings more and more, the DF varies more and more in an SET than a PP amp as the Ra is changing more in an SET triode while the PP triode combo is more constant. Theoretically, as one triode's Ra is rising, the other is lowering, maintaining an approx constant.
This Ra swing occurs in all triodes, with the Ra varying from several hundred ohms to thousands of ohms and eventually (theoretically) infinite ohms at triode cutoff.