This is an exemplary specimen of spin. The original research report considers many different sources of PM2.5 pollution and their contribution to additional deaths. The abstract then underline fossil fuel deaths as key finding, noting that they contribute ~27% of total deaths from PM2.5 pollution. The Arstechnica article just mentions fossil fuels and completely ignores the rest.
If anybody wonders, deaths attributed to burning biofuels (wood) are ~20% of total deaths (~800k), remaining are non-fuel sources (dust, fires, ...).