Yes it can, that's what I ended up writing at 4am this morning, lol. We manage way more instances then is feasible to do anything by hand. This is probably too late to help anyone, but you can also just stop instance, detach root, attach it to another instance, delete file(s), offline drive, detach, reattach to original instance, and then start instance. You need a "fixer" machine in the same AZ.
FWIW, I find the high-level overview more useful, because then I can write a script tailored to my situation. Between `bash`, `aws` CLI tool, and Powershell, it would be straightforward to programmatically apply this remedy.
