Requesters are the
only ones who can put exclusion quals on their studies. Unless I am mistaken, they have the solution right under their noses. Exclusion quals might not be perfect, but one requester in the comments on that site mentioned exclusion quals lessened the "bad data" to like 2%.
Sooo, they
have the tools to avoid this kind of problem but have instead resorted to blaming workers and using an arbitrary data criteria as an explanation for the 'bad data'. And last time I checked, 'bots' couldn't do surveys they weren't able to access