Does partial edge equilibration explain the $5/2$ thermal Hall value?

Test whether $\kappa_{xy} = (5/2) \kappa_0$ at $\nu = 5/2$ arises from an anti-Pfaffian edge whose thermal equilibration length $l_{\mathrm{eq}}$ exceeds the device edge length $L$. This scenario predicts $\kappa_{xy}$ depends on $L/l_{\mathrm{eq}}$ and crosses over toward $(3/2) \kappa_0$ for $L >> l_{\mathrm{eq}}$. An answer is $\kappa_{xy}(L)$ over a range of $L$ spanning $l_{\mathrm{eq}}$.

Thermal Hall measurements on $5/2$ devices with edge lengths varied by an order of magnitude at fixed temperature.