They take the average emission per kWh generated in the country, combine that with existing research of the social cost of carbon done by the EPA (which include

They take the average emission per kWh generated in the country, combine that with existing research of the social cost of carbon done by the EPA (which includes damages done through climate change), as well as the change in mortality rate caused by these emissions. Then they combine this by the electricity requirements of bitcoin mining, and they have their results. (in very global terms, there's a lot more work done in the results). I definitely wouldn't call it pseudo-science, as their arguments seem valid, and their methods seem to follow through at a cursory glance. They do note some weaknesses in their research, such as not having access to the exact sources of electricity miners use, however, which might change the results.