I would add #4: books that are at least 10 years old. I feel like knowledge has to pass the test of time. This rule would rule out "Why We Sleep" for instance, but you'll be able to read it in 10 years if the science behind it still stands.
Terence McKenna's suggestion was "if a book is less than 100 years old, don't read it".
I think it's a combination of "contemporary books will be from the swirl of life around you that already know, old books will be from a different culture and time and that difference is important" and "knowledge lost, from people who are no longer living".
It would also have the test of time - Darwinian natural selection of information. Something we're missing when we want the internet to keep data forever, we should be letting data rot and be lost, taking ongoing action to keep it in the present is a vote for its importance, setting up a system which preserves information without effort is cheating the system and leaving us swimming in e-waste.
I feel like every book is going to be contested both in the long and short terms. Instead of taking a book at face value, it is important to subscribe to the conversation around it that will inevitably follow.
This seems a bit silly. An analysis of British/EU relations that was written 10 years ago is going to be missing some pretty relevant things that have happened since then for instance.
