nothing in particular, but software is software. you know?
in analog, i think you can get plenty close with opamps, its just a part count/cost thing, and you lose some of the, i dunno, “grit”
and you can’t play your software mixer with a flashlight
sorry, but that’s the hand-waviest approach to modeling the vactrol that i can imagine. they just pop an integrator in there and modulate the integration time by the level a wee bit (like in any basic compressor plugin). then: “Adapting full models of photoconductor transient dynamics [25] for musical signal processing use is a significant task, and left for future work.”
most of the paper is about modelling the lowpass sallen-keyish circuit and the control path.
(and hey, nothing wrong with that, the simple model captures the broad behavior, just don’t go looking for too much special insight there)
in any case, i now feel extra lucky to have my little stash of VTL5C3.