It isn’t clear to me if the fundamental advantage of what I am describing is being communicated well enough, if that’s your idea. If you return to the project screen on your device and tap the Edit button, scroll to the bottom of your project list, and you will find a presumably empty section, headed by the phrase “On my iPad” or something to that effect. If you drag a project from Dropbox into that area then it is effectively offline. But note that when you tap back out of edit mode you can fully access this project. You can even edit it and later send it to another device. There is no functional difference between the projects in these two lists, save that one doesn’t clog up the sync mechanism.
As I understand it, you already have these projects stored in a different folder on your Mac, so they are available to it while you are there—the idea is to make them available to your iPad as well.
The way I would go about it is using iTunes or a better file manager to copy the contents of that folder into Scrivener’s document area with file sharing. Then you don’t have to waste the amount of time it takes to download them over the Internet. They should show up on your device the next time you view the project management screen.
The design intent of the software is to store only those projects you are actively working on in your sync folder—this isn’t a matter of opinion. The kind of sync we use is a heavy-duty process, necessitated by the types of tools we have available to us, and not at all like your iCloud Drive attachment in Files, or even the Dropbox.app itself. Those things have the luxury of keeping 99% of your stuff online and off your device. They also have the luxury of controlling both ends of the process, and thus being able to optimise how their server works with their client. We don’t have either luxury, and on account of how the Dropbox tools work, we must compile a full manifest of every single file and folder (which can be tens of thousands), each and every single time you press the sync button. It’s not ideal, but it’s the sort of technology that is available to this type of large-scale program.
So bearing those two facts in mind, it stands to reason that the more “dead weight” you store in this folder, the slower sync will be, and the more likely it will be that you run into technical limitations of all sorts, including outright failures when the context is suddenly dramatically changed around the stressed design limits.
Yup, I’d say it’s good practice in general as it will keep sync snappy, but until the bug is fixed, it’s a bit more along the lines of a necessary approach.
Unfortunately it won’t work for everyone—someone posted that the one WIP is too big all by itself to safely sync, but hopefully it helps some.