I found the PHS products useful for seeing the corridors of enhanced severe risk. They correctly showed that hail (or wind) was the highest risk compared to tornadoes.
ProbSevere was very useful with the trend lines. Again and again correctly identified risks as they occurred.
GLM showed strong ramp ups in activity prior to each severe issuance. No complaints here.
NUCAPS data was less useful, but with the skew-t pop up I was able to correctly see areas of mid level dry air. I did not use optical flow. Prob lightning was useful to see the first cells pop up but did not serve as much purpose after that with no known DSS opportunities.
Some Random Guy