Spinning off from T257597 and T254304 - I'm making a request here to confirm that all deduped contacts are being added to the suppression list.
Both @Eileenmcnaughton and I have noticed contacts in our database with data that doesn't make much sense, as seen in T254304. It's hard to query with a clear picture when many rogue contacts are pulled into the mix. When I run a query that I expect to return 0 contacts, I get a large return that warrants a closer look at the contacts case by case, which is why I cannot confirm when backfilled data from the civi export is successful.
One solution that might work is adding a way to filter out Master Suppression List contacts out of query searches, but we may need to consult with legal on that?
As this task is about data hygiene and Big English is just around the corner, I'm setting it as long term / low priority. I will drop in CIDs when applicable, especially if I find examples of contacts not on the suppression list with missing field values.