I might suggest that you use human hearing sensitivity to deal with your pressure test/simulations. Then, according to your aims, you can make structural modifications & tuning to empathise or damp spectral components. Start with an accurate and calibrated model (may be also hybrid for complex compounds) to predict the effect of your actions, then go to prototypes and full test sessions to certify you have got the desired target.

Best regards,

Alessandro

Hi

One way to deal with this subject is to make binarual recordings ('3D sound').

https://en.wikipedia.org/wiki/Binaural_recording

where you record and replay sound at the correct strenght and as unbiased as possible to a panel of listeners. This way you can compare the effect of versions A and B to see if anyone can tell the difference and if so, which is a good objective number for such sound.

There is state-of-the-art gear, e.g.

https://www.head-acoustics.de/eng/

and you can use DIY

https://www.youtube.com/watch?v=TH2b7m-UlmE

http://www.instructables.com/id/Binaural-Earbud-Microphone-AKA-holophonic-3d-ster/

https://www.musicworks.ca/diy/how-make-binaural-microphones

This kind of work usually is referred to as Sound Quality (SQ). SQ has over time developed lots of metrics, e.g. 

http://www.salford.ac.uk/computing-science-engineering/research/acoustics/psychoacoustics/sound-quality-making-products-sound-better/sound-quality-testing

Sincerely

Claes