Just adding the graphs from both the official GRAS rig with KB5000 pinna and the B&K 5128:
So once again, these should be treated as different ‘heads’. What you’re seeing here is how the headphone’s behavior changes from head to head. When placed normally, it seems fairly consistent across the two systems. So when placed as identically as possible, it does seem like the benefits of ultra low acoustic impedance are there.
Sadly, there is likely to be stronger variation across heads in practice, across different individuals depending on where the driver sits relative to the ear.