I've used that technique before to control the delay on a vocal, delay on a separate track, followed by compressor. Two sends from the vocal, one to the delay and one to the sidechain input of the compressor. So the delay really only comes into play at the end of the vocal where it swells up, very effective!
I learned that technique from one of his videos I sure. Great resource!
This is all beyond my knowledge/skill level but I do have a question. I often use automation lanes on a vocal track in Logic Pro to add or alter the amount of delay wherever desired or for that matter any of Nectar 2 modules and their many variables. How is this different in result from what you described? Pardon my ignorance.
Bud