if consensus can be reached regarding to a consistent and common set-up, I'd be most happy to weigh mine. I'd support a fluids full, powerhead only arrangment.
Wet-weight removes the obvious state-of-fill variables and reflects work-ready; while powerhead-only keeps bar-length preferences out of the mix.
I could go with full or empty and calculate this mass in or out, but having a hodgepodge of 1/4, 1/2, 3/4, 7/16 or whatever estimates really detracts from the validity of a dataset.
+1 How about full or empty, full would be better. Steve