Response to pressure changes is very simplified, all it is is partial-pressure difference that equalizes at ln(2) rate. The key parameter is time constant, specifically the time it takes delta-P to drop to half its initial value. You fine-tune the model by futzing with the number "tissue compartments" and their half-times, trying to back-fit your theoretical "compartments" to empirical observations (as
@Bernie_U posted).
You would typically have extra criteria to judge when you're "close enough". DSAT report, for example, mentions that 1990s USN tables were developed for 2% predicted DCS because the Navy considered 2% acceptable at the time.
I.e. you're actually looking at optimizing for multiple parameters, and the "real tissues" aren't even on the list.