One complete composed channel

For a preparation with several history blocks, the same calculation gives their subnormalized conditional blocks. It does not turn those history labels into the actual times of one fixed readout event.

At $\lambda =0$, Models 1 and 2 retain different gravitational preparations, so their complete matter outputs need not agree. What agrees exactly is the declared binary record law. No arbitrary later gravitational-sensitive measurement is included in that off-control claim.