I've simulated it and looked at measurements by people that have pursued it seriously. The Hannover hard-core guy has a great paper on it. Kemmler is correct, it is an interesting notion, and you can get it to sort of work, but it gets you nothing and creates a mess.
The distortion advantages...