I've been thinking about this after having read
https://www.nngroup.com/articles/participation-inequality/ again recently. That research was shared in 2006/9. They didn't quantify a upvote/reward driven website but I suspect hacker news is closer to the blog numbers than the general, so the percentage would be skewed. The closest statistics would be Reddit's, the self selecting "top 1%" subreddit has imposed a minimum of roughly 107000 combined karma/points which is roughly 114k users out of an estimated 52 million daily, 430 million monthly, and 1.5 billion registered total users.
You wouldn't need to crawl for your own dataset, it's available on Google's bigquery https://console.cloud.google.com/marketplace/details/y-combi... and slightly older 2006 to 2017 dataset is available on kaggle for direct download https://www.kaggle.com/datasets/hacker-news/hacker-news .
https://www.karmalb.com/user/rank:114000
https://www.reddit.com/r/top
https://earthweb.com/how-many-people-use-reddit/