You are correct in that as well as wrong values near solar noon the end of day values are in most cases wrong as well, although near sunrise/set the % calculations are not very good as the values are so low.
I’m not necessarily say Brian has deliberately changed anything, this could be the result of some change in the complier perhaps or just an accidental corruption in the code. Whatever it is this calculation in WD is now wrong. Until about version 145/6 it was fine and has been for years.
I believe the current version is 9.0 in build 149.
It makes no sense that this setup was working fine and one day just stopped. However, the folks at Weatherflow did make changes to the API (or something) because Brian made changes to cronweatherflow. I do not know what changes were made, but nonetheless, changes were made that could be affecting the OP.
Once data has been received by WD the code is the same assuming the station transmits w/m2 so the question for the op initially is can you see the solar values coming from the station, if none then you have a hardware issue. If solar values are coming in to WD then we need to see some screenshots of your solar control panels.
