Kalyan and I used a pulser fed into the PMT cables for a bottom-row shower PMT and a bottom-row cerenkov PMT to fake a T2 and T6 trigger with good timing.
These timings were measured at the TS:
BEFORE any changes: T6 arrived before T1 by 20ns. T6 arrived after T2 by 90ns.The timing between T2 and T6 was, at best, marginal with respect to the retiming circuit so we added 85ns to the T2 delay to compensate.
AFTER changes: T6 arrives before T1 by 20ns. T6 arrives before T2 by 5ns.