Hm, the author is conflating “reproducible” as in “reproducing bit-for-bit identical software” and “reproducible” as in “recreate the functionality of a operating system” which is what guix and nixos is currently accomplishing.
I don’t blame the author, it’s confusing when two closely related concepts uses the exact same word interchangeably. But it’s important to realize that nixos and guix doesn’t accomplishes bit-for-bit identical software any better than the old-school distributions.
I agree I didn’t write explicitly about the differences, except in the “verifiable” section were I discuss checksumming the system for possible attacks. Though in the case of the article, it doesn’t really matter as the meaning of reproducible is taken lightly. But for sure, I’ll be adding a footnote with your insightful remark. Thank you.