I'm fully supporting the position which states that using such an automated decision or bot will strongly decrease the quality of the bounty campaign - even self-educative platforms cannot fully analyze if the post(for example) is decent or it's just a spam. As the consequence - we will have lot's of "automated bounty" projects and tons of spam, even more than we have today.
However, bounty managers may vary as well, so the existence of such a bot may affect the "standarts" of a good bounty manager, that's pretty good for the community.
I doubt any serious campaign manager will leave it to full automation as the final decision in determining qualification. But the type of automation such as the one I posted above helps in eliminating human mistakes often made by even the best bounty managers. If you have a bot or tool that checks, among other things, for:
1. Minimum posts, post bursting, within specific cut off times
2. Unqualified posts, for example, in sections that don't count
3. Copy pasting/ plagiarism
4. Correct sig code
5. Correct application, including valid wallets, profile link
Then the human manager only needs to validate and check quality. You get increased quality, quicker validation, more time for manager to focus on improving participant quality.