I wanted to clarify that it is not 3% of all donors that go on to make an unreverted edit.  All donors are directed to a “Thank You” page after their donations, which includes several calls-to-action.  This year, we included a call to “Try editing Wikipedia”. Due to the way that we restrict tracking due to user privacy, the chart representing the editing funnel actually starts once a donor has clicked that call to create an account. In other words, of the people who donated during the Big English banner campaign, only 12,005 actually clicked through from the donor Thank You page to create an account. From those who click through, about 3.7% of people end up completing an unreverted edit. 

I see, that explains why this wasn't more than a footnote.  :)   Someone who just donated is unlikely to be a bot or spammer, so not surprising their edit quality is higher.

Once someone decides to try editing, ideally almost all of them would make an edit, without needing to make an account.  Perhaps: "Try editing" --> simple interface for easy edits --> make a few edits -->  see a few edits by others to review 👎👍 (or offer the same edits to a few people) --> only then make an account via simple creation popup...