Life-cycle assessment and carbon footprint studies are widely used by decision makers to identify climate change mitigation options and priorities at corporate and public levels. These applications, including the vast majority of emission accounting schemes and policy frameworks, traditionally quantify climate impacts of human activities by aggregating greenhouse gas emissions into the so-called CO2-equivalents using the 100-year Global Warming Potential (GWP100) as the default emission metric. The practice was established in the early nineties and has not been coupled with progresses in climate science, other than simply updating numerical values for GWP100. We review the key insights from the literature surrounding climate science that are at odds with existing climate impact methods and we identify possible improvement options. Issues with the existing approach lie in the use of a single metric that cannot represent the climate system complexity for all possible research and policy contexts, and in the default exclusion of near-term climate forcers such as aerosols or ozone precursors and changes in the Earth's energy balance associated with land cover changes. Failure to acknowledge the complexity of climate change drivers and the spatial and temporal heterogeneities of their climate system responses can lead to the deployment of suboptimal, and potentially even counterproductive, mitigation strategies. We argue for an active consideration of these aspects to bridge the gap between climate impact methods used in environmental impact analysis and climate science.