For your first point, I agree that the duplicates should be filtered out. I will schedule addressing that under this Jira.

However, I completely disagree with your second point (and your 3rd point relies on that as a prerequisite).