Because winters consume the most energy of any season (in the article) to heat, we emit the most c02 during that season.
Given that the planet is warming, the winters are milder (quoted above) (see this article on 13 of the hottest years on record have occurred in the last 15). Because winters are milder (because global warming), they require less energy to produce for heating.
The section I'm referring to is: ""However, CO2 emissions during January-March 2012 were low due to a combination of three factors: A mild winter that reduced household heating demand and therefore energy use"
Total heating in winter uses more energy than cooling in summer because there is a bigger difference between the outdoor temperature and the target temperature in winter than in summer. Just based on this article, one would expect the delta of warmer winters and warmer summers to roughly offset each other.
So it doesn't seem sound to conclude, based on this article, that, if the CO2 emissions were looked at over a full year, that the warmer winter would result in a net CO2 reduction.
That's assuming the overall US climate (by which I mean population- or energy-consumption-weighted degree days) necessitated similar amounts of winter heating and summer cooling.
Typically heating and cooling degree days (temperature variations integrated over time) are used to normalize energy consumption data for the weather. Not average temperatures.