Can phase information from vanished biological computations still select among symmetry-equivalent histories when the selection itself demands a broken invariance?