>> I want to be able to read it fifty years from now ...
> Yes, you won't be able to index them ...
The availability of old information is strongly related to the ability to index it. Your notes from ten years ago may be physically there, but how long is it going to take to find what you need?
I use pen and paper all the time but only for ephemera. Its advantage is that I'm not constrained by having to write pure text and I have a 2d surface, but once the thoughts are down and concrete the important bits are transcribed into my org-mode file, which even allows me to transcribe diagrams and math faithfully. I've read Altucher's words on using a waiter's pad and he says the same thing -- it's for ephemera. I've seen the way many people use physical notes and it's largely the same thing. The ones that do use paper as long-term storage have a separate set of notes just for that, and transcribe from short-term to long-term anyway, at which point the other liabilities of paper come into play.
Paper is heavy, and takes space. Over the years I've amassed a copious amount of notes that haven't yet been digitized. Without exaggerating, they're enough to fill a suitcase. That kind of mass is an anchor, and when I move I get to make a decision: bring them, or leave them in storage. Notes I can't get at might as well not exist.
My ideas, understanding of things also evolve over time, and if my notes are to be faithful to that they should allow the same mutability. Even if I want to preserve the history of the evolution, I'll want to be able to insert the new things where they make sense -- a new consequence here, a parenthetical that amends the previous thought there.
Paper worked well for persistence when that was all we had, but we have much better now.