The cruise limit on CHT was not that high. It was that high where it was measured--on the back side of the hottest cylinder, under the spark plug as a ring gasket. It was known that the CHT was actually much lower than that reading--by about 50dF. The CHTs were measured on only two of 18 cylinders, one in each row. When we put ceramic thermocouples on the cylinders, we found that the ring probe was reading higher by the referenced amount. The resulting operational CHT was in the range we now think optimal for longevity concerns. We know that cylinders operated under about 380-400dF retain a significant amount of metal strength over those operated at higher temps. We show that data in the new Master's Class given by APS.
The reason they advanced the timing in cruise was to place the thetaPP closer to optimal (16dATDC), while having a retarded timing for max power applications which allowed for a greater detonation margin and optimal crank-conrod geometry at that power. It was not about valve life. It had to do with optimizing the crank-conrod geometry for maximum torque on the given fuel burn LOP.
When running the TC-18 engines ROP, the TBO was 600 hours. When operating LOP, the TBO was 3600 hours. Why? Lower ICPs and CHTs resulted in greater longevity and fewer maintenance issues at the same power output. THAT was easier on the exhaust valves. They could not measure EGT on the engines as installed. That awaited the innovations of Alcor.