There's still no short summary of the method and demonstration that it works distributed.

Have you submitted such a report and had trouble publishing it, or is there simply no documentation that your forecasting techniques work?