OK, good points. We'll have to do a bit of investigation to see how far down the road we can go with hanging the runtime off of the Endpoints and EndpointReferences without updating the static model. I imagine that there is quite a lot of code that still expects it to be in the model still.
Maybe the approach could be to start laying out the algorithms for the component type and component configuration calculation. In parallel we can look at how/were to store the results of these algorithms, i.e. on both the model and the endpoints or just on the endpoints. Mike (and anyone else) all help is much appreciated. Regards Simon
