With Emacs org-babel I just do the refactor using the normal tools for whatever language on the tangled files and then detangle it back to the literate document. There's no problem here.
By "editing" I didn't mean editing the source. I meant the writing process, editing [0]. The thing you may hire an Editor to do. It is not something that you can automate, and the difficulty has very little to do with the tooling.
A slight change in code structure, may require many hundreds of adaptions to the literate document, for example.
I'd also say that a lot of the reason for refactoring is just... different. Literate programs are typically made to be fairly self contained works. If you have some general purpose code that is of use in the entire codebase, you can make that its own section and cover it when it is needed. Otherwise, you likely won't reach for the same coding strategies that are common outside of literate code.
I'm trying to find a way to describe this a bit better than the above. I think the easiest way to think about it, is that in most software projects you have a separate document that is the general architecture of the software. It is rare that you will need or want to refactor the architecture, so you try to keep that somewhat faithful to what the code is doing. In literate software, that high level architecture view is part of how you organize the code.
It's simply more work - but that "more work" is vitally important, tedious, and resistant to any kind of automated help.