Towards a More Complete Theory of Function Preserving Transforms