Growing up in Ivory Coast, the weather bulletin after the evening news was a lot more about how the weather had been that day.
I should dig out a spreadsheet I worked on tracking weather predictions daily for like two months. Took the rolling ten day forecast and measured change and accuracy. I'm an exciting person, i know.
I keep meaning to do that, but I have more important shit to do... like post on here. I do notice that the 7-10 days out forecast often calls for astronomical amounts of rain, by the time it comes closer, 2-5 days out, it's only torrential rain; when it's the next day's forecast, it's a pretty good amount of rain... then on the day in question it becomes a 'chance of rain' which sometimes actually happens. Maybe I will start tracking just the rain forecast vs actual for 7, 8, 9, and 10 days out. DFW would be a rainforest if the forecasts held true!
Rain varies a lot by microclimate and storm patterns. If you were to do that it makes most sense to take a lot of readings and area average them, because unless you call the office for a spot forecast (which they typically wont do for more that 24 -72 hours), what you are getting is a regional/psa one. Mesonet is a good archive.