Noslop -- Thanks for the witty presentation of your empirical findings! :)
But I don't think that they necessarily support the conclusion that the rca-to-xlr conversion is what is responsible for the huge volume difference that you heard between the two configurations.
Even if the power amplifier architecture corresponds to what Rwwear had suggested, which is the worst case in terms of the possible effects of the conversion on the amp's power output, there would only be a 6db loss in maximum output power. That represents a 75% reduction in maximum amp output power, which in subjective terms is not anywhere near the difference you describe, imo.
Putting db losses into perspective, it is commonly said that a 10db loss (which is a loss of 90% of the maximum output power) corresponds to a subjective perception of "half as loud." What you are describing sounds like a greater loss than even that, given that with the balanced preamp you had the volume control at less than 1/2 scale, while with the unbalanced one you had it at max.
So something else is going on, obviously related to the extra variable of having two different preamps in the comparison. Perhaps their gains are different, or perhaps something in the complex set of settings of the Stage One is confusing things.
I think that a better test would be to compare balanced vs. unbalanced outputs of the Aurum, with the unbalanced Aurum output going into the power amp through the adapter.
To summarize what has been discussed as to the different possible effects of the adapter on the power amplifier output, depending on its architecture:
If the power amp architecture is as I and Atmasphere envisioned, there would be no sacrifice in maximum output power by feeding it single ended through the adapter; it would just be necessary to raise the volume control 6db higher to reach that maximum output power (compared to balanced drive).
If the power amp architecture is as Rwwear envisioned, then the maximum output power of the amplifier (the point at which it would clip) would be reduced by 6db (compared to balanced drive), and that reduced clipping point could not be overcome by the volume control.
I think that the fact that you were apparently unable to clip the power amplifier when driving it with the adapter, even with the volume control at max, reinforces my view that your findings so far are inconclusive.
Regards,
-- Al