I would suggest naming a range of valid hardware platforms, or that platform used to benchmark should be a part of each submission. Otherwise you might waste time discovering that someone was using some underpowered embedded device (or a Nintendo Gameboy) to produce their quirky, but totally honest, result
(Added your suggestion to OP)
I'll also try to publish a final benchmark in a comparing way with a regularly used script as the base-line and anything else I can come up with or find here as the rest.