Nick, now I see the error in your reasoning. If you look at the Hugo¹ and Hugo² specs, there's no logarithmic increase of output power with a logarithmic decrease of impedance. Because the current flow in common amps is limited. The DAVE comes closer to it, even achieves a perfect congruence between 300 and 33 Ω: ¹/909 of the impedance leads to the 9,09-fold power. However, with even lower impedances it certainly won't behave like that anymore.