But they are good at processing the current day's weather, and drawing simple patterns based on notable weather anomalies. So that's exactly what they do.
This was one of the techniques encouraged at a talk I attended a year or so ago which was aimed at getting NASA scientists to better interact with a non-scientific audience and media. If I remember correctly, the average person hears that the scientist's data/trend is statistically significant then they take that 95% confidence as being "5% chance that it's all wrong." In all probability, they won't take steps to learn why this is incorrect no matter how many times it's explained. In this case it's helpful to encourage Joe Sixpack to make a mental note of days with a record high or low. In this way he can see for himself how his local climate is shifting and in which direction.