These graphs really would benefit from error bars, especially the breakdowns into sector. I would not want a policy maker to rely on those numbers.
It's a constant battle. A politician with a science report has official CYA for whatever they do, as they were making the decision based on scientific estimates.
Assuming they're not broader than the signal being measured, and they certainly won't be for recent history, it ought to actually make it more useful for politicians
Politicians always push for a scientific consensus they can act on, whether it's there or not.