We are currently managing about 500+ workbenches, some in the filesystem some on FME Server. We try to name them based on the datasets they are working on and the source and destination systems/databases.
We do not know if it´s the best way to keep up with finding them again, but up to this scale it works.
A best-practice HowTo or tutorials about workbench documentation and management at larger scale would be great, maybe there are better ways than we do.
This is how I usually manage workspaces on a file system:
/<project>/<theme>/<verb>_<dataset>.fmw
Of course all the strings are sanititzed (i.e. no weird characters etc in the path names) and everything is versioned with git.
On the server I try to use the the same methodology. Since you cannot nest repositories, I concatenate the repository names as <project>_<theme>, whenever necessary.
Hopefully that can give you some ideas.
In addition to @david_r's comments, the last part of the Desktop Basic course covers best practices. It does focus on the desktop / workspace development aspect of it and not so much on a more overarching management of workspaces, but it's a start.
I would also recommend, if you haven't done so already, to standardize the workspace layouts, for example by using a template such as this one by @lars_de_vries. This will help you maintain the workspaces with a larger group of people.