I was trying not to make this into another thread on my favourite subject (honestly! ), however, I agree with this advice to the OP. Since I read what you wrote the other day about the torque multiplier in particular I’ve been using the small ring far more in workouts and you’re right that in some circumstances this seems to confound my theory a bit!
Having said that, and back on this topic, about the OP’s potential ramp test variance, I still can’t get away from the fact that at, say, a steady 95rpm, 250W, ERG mode, in 53-15, my RPE and HR are significantly lower than the exact same 95RPM/250W if I ride in 39-15, and even small reductions in gear ratio produce a detectable increase in RPE and HR, albeit less so.
My caveat is that it may just be that the Kickr Snap (which I use) makes this phenomenon more pronounced than in a direct drive, perhaps because its flywheel is driven by the fast-moving surface of the tyre on a very small diameter roller that’s directly attached to the flywheel (a big gearing-up effect). I’m not sure how the direct drive Kickr flywheel is geared from the speed of the chain at the rear cog to the flywheel - is it direct or is it via a toothed belt and sprockets? If it’s direct then the flywheel speed would be far slower at a given cadence/gear ratio than the Snap which may explain the difference in the flywheel effect on gear selection when comparing wheel-off to wheel-on trainers.
I’m sorry, I think I’m starting to disappear up my own a***e, but it’s a genuinely interesting subject!