I just found this[1] which mentions rubber vs. foam surrounds (suspensions), and butyl rubber with apparent good life. (That might have been the simple difference; the normal-looking speaker was already sourced; I came in to do embedded/DSP.) At kilowatt/constant power levels I'd imagine going as far as oil/magnetic suspension for high life; even o-rings under high cycles can fall apart.

I'm pondering how one would do accelerated reliability/life testing for the acoustic heat pump application. Maybe they have good-enough FEA/multiphysics on a suspension and cone, and strain/life curves they are prepared to trust. With a slope on log-log strain vs life they might project how much to overpower/overstrain for an e.g. 1000hr accelerated test.

[1] https://www.avforums.com/threads/speaker-life-expectancy.164...