I mean, so is Pascal's Wager. I'm no climate scientist, but from what little I've read is that _all_ climate models are terrible at predicting more than 10-20 days out, and are on the same level of accuracy as financial forecasting models (i.e., 'not at all'). In my vast and inky ignorance, I'm just not convinced we need to _drastically revamp the entire economy soup to nuts_ in order to satisfy predictions from inaccurate models populated with bad (or at least 'wildly variable') data.
> climate models are terrible at predicting more than 10-20 days out
Note that if your weeding is in 30 days at noon, and you want to know if it's going to rain, then you are doomed. There are no good enough models for that precision.
But if you want to estimate the total rainfall during the year the problem gets easier. Not very easy or easy, just easier than predicting which day will rain in approximately a month.