Update GMMs that use delegates to get models from cache

Currently models like AtkinsonMacias09 that delegate to other GMMs commonly create dedicated instances of the delegate as opposed to getting the model from the loadingCache and casting back to the concrete implementation class.