Dear Blick, Felix, and WeatherCat faithful,
I think Phil Gladstone's system still needs tweaking.
I agree 100% with your basic points, but I fear you might be tilting at a windmill on this one. I don't know if there is a well-established methodology to rate the accuracy of weather data from station to station. If Phil Gladstone was forced to come up with something that is somewhat arbitrary, there simply might not be anything easily available to create ratings that are demonstrably better.
I have no idea who uses CWOP data anyway. I do know a bit about using data where the reliability of the data isn't certain and you can use statistical tools to make inferences even if you cannot trust all the data. So researchers who need a shotgun approach to data can simply use it all without worrying about station ratings. If researchers do really care about data from a particular station, I think they would take the time to check if the station is setup correctly.
So I don't think the research community is likely to pay any attention to Phil Gladstone ratings. That only leaves us amateurs. It may hurt our feelings a bit when our stations are unfairly derided. But if clearly the rating system is flawed, we should probably just let it slide and move on to problems we have some hope of addressing.
I doesn't make me feel better either, but these days I need to be choosy of the battles I decide to fight.
Oh well, . . . . . Edouard