I took a peek. If you worked here you could actually run your model against a month of our data and see for yourself what you think of the results.

But just at a high level, there are two things that I think you are not factoring in. One is that the multipliers in our payment far outweigh the engagement signals like claps. I think people who are focused on claps and highlights are missing the point.

Then the other is that about 2/3 of technical writers don't write behind the paywall at all. There's a certain payment level in partner program where the quality is equal to or worse than what people are already writing for free. So that's one of many reasons it doesn't make sense to spend reader's money that way.