New here. I'm interested since I know Charles well and have a high opinion. Same for John. It is certainly possible that that unit was defective. I'm wondering if the dip switch was maybe defective. If the actual response was set for one of the boundary setups, instead of the assumed no-boundary setting, that could shave off 2 dB at the bottom, which kind of matches the deficit in the measurements. I'm not sure the exact filter cutoff for those boundary correction filters, but it should be a few hundred Hz.
I am also familiar with the Klippel system, though I have not used it. I did take the Klippel seminar, it was like drinking from a fire hose! The NFS does have diagnostics to check fitting error, and maybe suggest more data points or higher order wave expansion if needed. One thing is the selection of the acoustic center (expansion point). My Klippel manual says it should be the center of the tweeter, but the accompanying picture shows it between the tweet and woofer, like many people use for normal chamber measurements. Could that have any bearing? I assume the ugliness at 8-10 kHz has to do with the coaxial design of the mid/tweet? Diffraction.