So they intercept the IR. The point is they have a small effect since other molecules, like moisture in the air, do the same thing and the CO2 molecules are vastly outnumbered by them.
In most cases. But in some cases the CO2 intercepts different frequency IR since CO2 and H2O vibrate at different frequencies. Also there are cases where there is 400 ppm of CO2 and 20 ppm of H2O since the latter varies from 20 to 20,000 ppm. The interesting thing is that the variation in H2O is simply due to weather and thus weather is what actually controls the equilibrium temperature of the planet which can jump up or down by 0.2C in a few weeks (equating to putative warming of decades of manmade rises in CO2).
In an attempt to prove that CO2 matters they will often point to the Arctic which is dry and thus CO2 should matter more. But while the Arctic rose by almost 0.5C per decade for a while, it has lately leveled off (6-7 years) trailing the lull in warming planet-wide. That implies that the Arctic follows not leads. They never point to the Antarctic which is even drier on average unless it is to use the funny number studies like Steig et al which smeared a bit of local ground station warming across large swaths of W. Antarctica.
Your vastly outnumbered premise is correct on average, but not always, so it is worth considering alternatives. Unfortunately the only way to test the alternatives is the worldwide climate models which do not model weather and like I said, average weather is the key factor in the variation the planet's average temperature.