This doesn't seem right -- it wouldn't detect multiples if your code below is correct (please add some multiples as tests, if possible)

Extended regex to include search page.

let is prefered over var for loops and similar temp uses

It seems like you're just generating a crude version of ZU.cleanAuthor here -- I'd recommend using that.

I missed that function. Replaced the logic I had written, but some preprocessing was still required due to how the byline string can be formatted.

use attr and CSS selector here instead for simplicity

same here

(\d{4})(\d{2})(\d{2}) is more precise and more readable

remove all TODOs

using Zotero's built in function will solve this, but the right way to handle single-name and institutional authors is to use lastName only and set fieldMode: true

These tags look generic and not helpful. I'd clean them out.

I wasn't sure what tags did, so I kept them as they were. As you say, at least some of them are quite general and don't seem to relate to the article. Happy to get rid of them.