Proxima Fusion interview question

How does a higher order lower resolution discretization trade off against the current implementation