I would say that the best strategy of any backup is to start backup things asap.
Start with the easiest option and backup everything into one repo. Restic will handle it without any issues IMO. And only if you discover that it does not work for whatever reason, think about complicating it:) Unless it is some huge organisation server with thousands of users/millions of emails and hundreds of gigabytes of databases data there is no point really to over engineer things.
Tagging different content is pretty good idea to keep things organised. Also handy to apply easily various retention policies for different data types if needed.
Different retention policies for different content types
Maybe you want to rsync repos only for certain content types somewhere else (cloud?)
Possibility to later use separate locations or access permissions for different content types
Main reason for using a single repo imho: deduplication (if data chunks reappear across content types). This works extremely well and saves me a ton of space in nearly every repo I’ve ever created. As @kapitainsky said, restic doesn’t care about things like file number or size. No restrictions here until you get to a certain point.
