Sergey,
That is very on point. In a broad based statistical analysis, how do you separate the “signal” from the “noise”?
When a discipline or study is very focused and controlled, knowledge can be extrapolated and applied.

Contrary, in such a broad system as the planet’s meteorology, where do we draw the bounds of what is a proper input?
Dust? greenhouse gases? Volcanic debris and dust? Cosmic rays? Solar Wind? Sun spots? Geologic activity that could be warming the ocean locally, or even planet wide (think of the Ring of Fire in the Pacific)? The flapping of butterfly wings (OK, that’s just silly)? The flapping of my gums (sillier still)?

So many inputs, so little control. Greenhouse warming may be affected by man-made chemicals, but it seems so hard to separate out effects and weight them carefully. CO2 is implicated in global warming, but atmospheric water vapor is a stronger “greenhouse gas” than CO2. Should we then control water vapor?