To your first question - By all means, you could prove it. Establish relationships with and systematically monitor the displayed and filtered reviews of a statistically significant sample of businesses. Record when they are contacted by Yelp sales reps, record their decisions on those calls, and classify their filtered review behavior before and after. This is doable, and it would be enough information to determine if there is actually a correlation in review filter activity to advertising offers.

Go for it. As a former Yelp engineer, I can assure you that the answer would be negative (no correlation), but I invite you to give it a shot.

To your second question, the answer is "yes". A spam fighter simply cannot publicly release their spam detection algorithm - it would make the system too easy to bypass. SEO black hats have been trying for a decade to get around Google's sorting algorithms, and then they cry foul when Google changes things up. The message from Google all along has been: Stop trying to game our system, and instead focus on creating good websites. Yelp's situation is no different.