On Feb. 27, while investigating some strange scaler rates, I disconnected T4 and reconnected it incorrectly. Instead of being the correct logic, the output of the MLU was essentially S1 AND S2.
This was corrected on Mar. 4.