I'm a professor and I write a lot of lectures using RMarkdown and RevealJS.
One annoyance that I have is that knitting a file involves either 1) Having a ton of revealjs files in the directory with my HTML file or 2) Creating a standalone file that takes a while to generate and is unnecessarily large. Option 1 isn't terrible but it has some downsides, e.g., cluttering up my git history and making it tough to track important changes.
yes Quarto works a bit differently that rmarkdown and there is not lib_dir configuration. Managing paths is complicated and prone to errors so everything is currently located right next to the files. However, lib_dir is a output type of folder, not a source type. Using Quarto projects, you could have specific output directory to store everything.
I understand from the above that you are tracking on git your output and not just your inputs ? is that right ? Usually the source should be enough to reproduce and track on git. Especially using the freeze feature that allow to quickly reproduce an output without any recompute if nothing changed. (see Quarto – Managing Execution)
If none exist yet, you could open an issue in Quarto repo to discuss a solution.
Only option that could help for revealjs is reveal-js-url to change the revealjs version used: Quarto – Revealjs Options
I see. Quarto has a way to publish to gh-pages easily: Quarto – GitHub Pages. Usually duplication of resources is not something I look too much as it leaves in another branch where I published directly without checking change - I am trusting review in source.
Anyhow, possibly reveal-js-url will help you here;
Do no hesitate to open a feature request if something is missing for your workflow.