Conclusions
There is no one way to verify a forecast.
Verification is needed to calibrate forecasts within and out of the NWS
The accuracy of light and moderate thresholds of precipitation has increased
for the highest threshold there has not been a commensurate improvement in forecasts (at least that is what deterministic verification suggests).
New and innovative ways are needed to verify forecasts of small scale events and high resolution models.
I’ll be retired, dead and buried before we really get good at it.