Took me way-too-long to get my head around the idea (elaborated above) that if you took 6 various headphones and measured them on all the different measurement rigs, that you don't get a consistent delta between any two rigs -- it is headphone/rig, geometry/wavefront dependent. That is, it's impossible to create a single compensation curve to translate marv->tyll measurements for example, since the delta varies from headphone to headphone and the spread can be quite large.