Earlier today, I posted two experimental computer model forecasts valid at 7 and 9pm this evening. How did they do? In Iowa and Nebraska, very well. In Kansas and Oklahoma, poorly.
7pm Actual Radar:
9pm northern forecast and actual radar:
This is a very good performance. Keep in mind that you cannot forecast thunderstorm locations down to a county.
9pm actual and radar, southern forecast
This forecast was lousy. The “cap” (an area of warm air about 9,000 ft. above the ground) apparently kept storms from developing in Oklahoma and in most of Kansas. While the one storm near Lyons did cause a trail of large hail, it doesn’t make up for the lack of thunderstorms compared to the forecast.
So, there is promise shown by thunderstorm-scale computer models but they still have a long way to go.