Glad to hear that you have fun with it.
Certainly different distributions make a difference, consider: not only for dur but als for amp and here you multiply the Dbrowns in addition. The built-in distributions make the Gendy-UGens comfortable and close to Xenakis’ definition. But on the other hand with DemandEnvGen you have much freedom with defining arbitrary sequences, also you can sequence curvature, as often a trade-off …