I think you’ve put your finger on the difference between Berkeley’s Richard Muller and me. He seems to feel as you do that climate is BEST understood through millions if not billions of temperature datapoints.
I’m way at the opposite end. If you want to make projections about Earth’s climate 30 to 100 years into the future, I claim one can do a very good job by recording just one temperature measurement per year, and that a billion measurements a year tells us little or nothing more about our long term prospects than just that one measurement.
Though not if you start today. Starting around 1850 suffices.
Furthermore I claim that with that one measurement per year one can learn things about the climate that weren’t even suspected before, things that you wouldn’t have noticed had you been burdened with a million or a billion datapoints a year.
But that’s just a personal belief of mine, and I respect that you take Professor Muller’s side on that question.