I read in an article on modular synthesis that "wet" sounds come from different frequencies at a cascade of different delays. I tried it out using Reason's vocoder to split up the spectrum once, hooked each stream up to a increasingly short delays, and it more or less worked. There must be easier ways to do it though...