The problem is that this oversimplification is wrong. Curvature in general relativity cannot be expressed merely with a 2-d matrix. It's actually a hypermatrix a