Databases are separate entities though you can construct ones that have logically connected (or even redundant) information in them. All your separate databases do not constitute a singular body of work, except perhaps conceptually.
Databases are not merely groups of data. They are more akin to books. If you have two books about whale watching (and who doesn’t! 
), each book will have its own glossary of terms even though many of the words would be the same. If you wanted to use one set of terms, it would be better to create a compendium, a massive volume of whale watching information, with one set of words in the glossary.
Per this analogy, yes it is possible to create a larger database with one set of tags. And if you had a ton of related information, I would suggest keeping those things in one database.
What if I want to clean up my tag list (which I do from time to time)? Then it is very long and confusing.
You could use @cgrunenberg’s smart rule to clean out empty tags. Let me look for it, in case he doesn’t chime in…
https://devontech-discourse.s3.dualstack.us-east-1.amazonaws.com/uploads/original/2X/7/793544bd1df66ba6c711fd38116311f78922bd74.png
From: