Or perhaps a "scientist" has spotted a suspected cancer in your brain but they can't quite pinpoint whether it's a 20x increase from normal size or a 28x increase. Based on the 28 assumption, you'll be in a wheelchair within seven years, but it might also be ten years. It's entirely preventable by doing the surgery but that costs money today plus minor lifestyle changes. Before doing a costly operation and bothering with, say, a daily pill and no alcohol, one could choose to wait while they refine their equipment over the next decades and then measure again before taking serious action. Can't trust dem scientists anyway, maybe it wasn't a problem in the first place.

That's the logic I'm seeing here. A "times worse than CO2" value for one of the elements might have been off by 30% so we should just assume nothing is wrong in the first place and not change our society away from burning carbon. Maybe it's just a coincidence that ~five of the past ten summers are the hottest of all the summers whose temperature we've measured... yeah I agree: maybe. It's a possibility, no matter how remote. But I'd rather not wager with billions of lives impacted by the outcomes that are part obvious (sea level rise) and part the prediction from yet more models. We needed to act yesterday if the models are anywhere near the truth and it's going to get exponentially harder to fix the longer we wait (because reducing emissions year-on-year is a lot easier than cutting+capturing at a moment's notice). If it turns out to be a dud, we can continue to dig up the remaining oil and have a big feast, everyone happy. If not, it's a good thing we acted and we're probably still not going to be able to prevent some level of habitat damage for the species we call human, but at least it won't be as bad as when we wait for more precise information on a problem where the conclusion remains unchanged.

(That's besides the healthy years of life provided by having cleaner air; I don't know how that would compare against e.g. cobalt mining.)

And that argument works better in reverse anyway: we're talking about a trend implying large parts of the planet becoming uninhabitable for present levels of human population within a relatively short time frame (which will also have an effect on economic activity!). If contrarians really have a good faith belief that this trend is caused by some other factor and will be self correcting, perhaps they could deign to specify their own testable assumptions and make their own model...

I'm not sure what sort of model you used to arrive at this vague estimate, but:

Even if you assume AGW science is completely wrong and the effect is zero, humans have to move from fossil fuels to renewables/nuclear in that timeframe anyway, because we have 47 years of oil left. Whether it's 20 years or 80 years, still needs to happen.