Apparently the way it is presented is that, because scraping per definition alters the content of the copy (i.e. it extracts parts of it), it is therefor a modification of the original works. This is mostly the problematic part.

I have thought about just using the DOM, and/or using Chrome itself as an intermediary using its dev toolkit but I think that from a legal perspective, this is all just a modification of the original works as well.

IANAL, of course.

I'm pretty sure that's not the way it was presented at all. Modifying original works is actually an allowance of copyright, not a restriction. It's called Fair Use, which states that "transformative" uses of copyrighted material, or "those that add something new, with a further purpose or different character, and do not substitute for the original use of the work," are "more likely to be considered fair." [1]

Extracting the user's data out of the page is what's _allowed_; since that is the part of the page that Facebook does not actually hold a copyright claim to, I'm pretty sure you could make the claim that extracting the user's own data out of the page into a structured format for the user's own use is transformative.

The part that wasn't allowed was literally making an ephemeral in-memory copy of the page as-is, which included Facebook's copyrighted HTML.

(I am also not a lawyer)

[1] https://www.copyright.gov/fair-use/more-info.html