I actually also have been curious to see if the amp can measurably influence the measured frequency response. I had a LLmk2 in my possession for a short while but was dumb enough to not make any measurement. I could however compare the 727A to the BHSE if I ever receive it
. That is assuming I can have sufficiently repeatability (a big IF
Having said that, from the few simulations I ran as linked above (mind you, they're based on guessed material properties so only worth so much), the sealing is indeed the main driver for all this.
By physics, a sealed baffle plate headphone in a sealed earcavity (e.g. Leather/pleather pad), should not roll off the bass. If so, there's got to be leakage somewhere.
Now, the resonance you see (typically a blip at 30-70Hz depending on the phone) is combination of diaphragm / air cavity / earpad cushion effective mass & stiffness.
Coming back to the lambdas, they thus have the potential to measure flat if the seal is right, just like the omegas and other 009.
Also, as for Tyll's stat measurements, I thought he did them all in one or two scoops with the help of the "mafia"? E.g got a bhse lent to him and phones from here and there. I might be wrong though.