Is it really that hard to forecast the amount of power residential PV generates? I'm pretty sure you could build a model that is pretty good just based on weath

Is it really that hard to forecast the amount of power residential PV generates? I'm pretty sure you could build a model that is pretty good just based on weather data, but failing that why not just put a single small PV panel on each (plant, substation, power pole... choose your level of granularity) and have it report back the amount of solar energy in that location? That should get you a very good idea of the amount of power coming in.