I don't get how the no-hiding theorem implies that the information will be preserved in the remaining qbits, if the "environment" of the measurement is the lab, not the available ancilla.