Follow-up from "Change BitBucket Pull Request Import to a Batch Process"

The following discussion from !46696 (merged) should be addressed:

• @s.schrottner started a discussion: (+1 comment)

  i understand why this is here, and added, it is also the reason why it has a batch size of 100 - as the comment is stating - but are there any statistics on how much benefit this adds, performance wise, or if we should see this as an extra step, and remove from this method (the cleaning is also a detached method)

  i still think that this mixes some concerns, and i do not think we should bother about calling an api for an import to often - as the import will be done normally just once per repo