he infrared absorption of carbon dioxide is experimentally measured in the laboratory
No one rational doubts this. That has never been what the climate change debate was about. But the atmosphere is not a bottle of air, or even a bottle of air and water (any modern meteorological model treats modeling he ocean at least as importantly as modeling the air). The atmosphere+hydrosphere is a complex, evolving system with many feedback mechanisms, both positive and negative.
I mean, really, do you think a climate model is simply modeling a static stack of air with some CO2 in it? Really?
The question is: quantitatively, what rate of human CO2 emission with create what effects, in detail. This is not the sort of science that lends itself to reproducible experiments, but that's fine, neither does astronomy or cosmology. It is, like any science, required to make falsifiable quantitative predictions.
And, frankly, the best models aren't doing so well, giving about 2 sigmas of accuracy. If you generated hundreds of models at random, you'd expect a couple dozen to have 2 sigmas of accuracy. That doesn't mean the models are flawed in any fundamental way, but there's a big gap between "not fundamentally flawed" and "great, proven science".