The source on Embodied Carbon in SSDs seems to be this 2022 paper: https://hotcarbon.org/assets/2022/pdf/hotcarbon22-tannu.pdf ; but unless I'm mistaken, it does not claim anything like the full Seagate "report": https://www.seagate.com/content/dam/seagate/assets/resources... ; there might have been an "extrapolation" because 30TB SSDs did not exist then (outside labs).
I'd treat it with a two-order of magnitudes (at least) grain of salt.