I'm not sure that invalidates the reasoning behind nuking hashtags driven by bot farms, which is clearly a practice that negatively impacts the practice. What you're describing sounds like a case for better discerning between bots and real users.