Does anyone know of any sources that attempt to look at the marginal energy source rather than the average energy source?

For example, a region might have 50% of its power from hydro and 50% from gas, but 0% of new power plants are hydro and 100% are gas. In such a world, adding a new electricity consumer to the grid causes 100% gas emissions, even though "50% of power is clean hydro." The proper attribution really ought to be the causal attribution, which typically means looking at the marginal source of power rather than the average.

I know this is a complicated topic. Entire PhD theses have been written on modeling the emissions impact of EVs, and there's a healthy debate of how to best calculate the 'causal' impact.

as long as there are any dirty megawatts, any clean bitcoin mining energy could be replacing it, assuming it is physically close enough to existing electric power grids.

all bitcoin mines are is a proof of investment for the purposes of divvying up entries in a decentralized periodic lottery where the probability of winning is proportional to ones investment in equipment and energy.

is it really truly impossible to efficiently simulate this (specifically the economics, so that existing players don't revolt) while maintaining decentralization in the consensus algorithm?

it seems like there could be a technical solution, but the harder problem is building political consensus in a decentralized world... sort of like getting the global economy off of fossil fuels...