That “rant” of mine doesn’t use a 20-year dataset. It uses five twenty-year datasets, every one of which shows a trend which has clearly emerged from the noise. Not only that, but a trend which is increasing with altitude in exactly the way that’s expected from understanding how the greenhouse effect works. Statistics 101 (well... maybe 201): the ratio of the trend to the magnitude of the noise controls the amount of data you need to reliably see the trend.
