I don't have time to search for the exact reference or the numbers, but there was a European study of the total life cycle environmental costs, including CO2 and other pollutants, of various energy technologies. In terms of CO2, hydro was lowest. Nuclear, solar, and wind were roughly the same. (I believe nuclear was computed two ways, once with gaseous diffusion--still used, but being phased out--for enrichment and once with gas centrifuge. Gas centrifuge produced lower CO2 emissions, but neither figure was astoundingly high.) I believe nuclear (and wind and solar) come in at around 6% of coal. The concrete and steel use in a nuclear plant was taken into account in the study, as were emissions from mining. There is also a Swedish environmental report with similar conclusions.
As for what percentage CO2 reduction in US power plant emissions could be expected, that would depend on how much of the new capacity replaced gas, and how much replaced coal. Any (expensive) gas used for base load would be the first candidate for replacement. That would reduce the impact on CO2 because baseload gas fired plants, expecially the combined cycle plants most useful for base load generation produce less (half?) the CO2 per unit of electricity generated of coal plants, though still far more than nuclear plants. Once you cut into coal, which produces about 50% of US electric power, you see some serious CO2 savings.
Two other comments:
The statement was that there would be a 30% reduction in emissions from POWER plants. I think electric power plants only account for about half on US CO2 emissions, so you would only get a 15% or so reduction in overall CO2 emissions.
From your question, I suspect you are being misled by a discredited Dutch study which claims ridiculously high CO2 emissions for both nuclear plant construction and uranium mining.