Well, Picard is mostly relying on data added to MusicBrainz, you should rethink your workflow so your work benefits to the whole community and you get most benefits of using MusicBrainz:
- search for artists and releases in MB database (you can use Picard for that too)
- check if existing metadata matches what you have
- edit or add data, including cover art, links, tags
- open Picard, select your files, select correct release, save them along retrieved metadata
- you can also add Acoustid fingerprints, ISRCs, etc…
The main benefit here is that Picard will store MBIDs to your music files, and you can retag once in a while to get fresh metadata.
Another benefit: your metadata is shared, online, spread over the internet, it is more unlikely to disappear, compared to be only stored in your own files on your own storage.
Let’s say you add a track titled “My Fancy Tilte” doing a typo, someone else may notice it and fix it, next time you re-tag Picard will retrieve updated title “My Fancy Title” and matching file will be updated (file name and tags).
That’s the whole point of a community-maintained database.