Not sure if you have had a chance to go through the wonderful models and theories behind the impulse propulsion energy system.
They are prototyping a PRD1 which is a hand crank started model that is slated to produce 166% coe. We'll see from the physical testing if that is proven or not. What I am wondering is, why doesn't someone use a decent electro magnet and some leverage to repel the stationary emagnet to move the handle down to a repulsed emagnet that will move a lever to produce the required start stroke of the process? How much energy does a small emagnet need if properly leveraged and/or weighted? Think of a pair of bolt cutters as the leverage mechanism possibly.
Here is the link to the prd1 http://www.planetenergy.com/pri10.htm
Attached is a picture of what I am talking about to make it a self runner. Also they are fully intent on making the PRD2 a self runner by having a small electric motor to do this. Think about replacing the springs in the whole mechanism with electromagnets that are accurately pulsed  on/off and possibly using two e magnets so that one is the repel s to s and one is the attract s to n to replace the springs. Also, they should be considering magnetic bearings on all the axles also. That would reduce the maint costs and downtime. What about replacing all the brackets and other possible parts with hard plastic compounds or carbon fiber to make it lighter? I am thinking car, plane, boat sized to run.
Thanks.
BE