Even if that might be the case now, I doubt that holds if piracy becomes truly widespread.
I would suspect A pirates book B and tells C about it, C buys book B is a lot more common than A pirates book B and likes it enough to buy it
I have no data to support this, and while I have paid for things I could access for free, but I'm sufficiently pessimistic about human nature to think that's the norm.
Piracy has been "truly widespread" for decades now.
Most people who are able to, still pay for things, especially if they're convenient. Even when those services actually add additional restrictions to their access to the media they think they're paying for.
"Despite the data, I continue to guess that in the future my hypothesis will be true."
Bold. Not inherently incorrect, but not optimally heuristic.
Despite which data? https://news.ycombinator.com/item?id=15305476