I have no idea how spinning up a MicroVM solves the problems this paper talks about and I doubt you actually have programmed GPU kernels ever

I just read this tool up and down but still can't conceive of a use case.

No, it’s not relevant to simple GPU offload, stop spam linking your slop blog.