That's the average. Some brain cells don't churn as fast (or maybe at all?). The memory-of-a-memory thing isn't on a 7 year scale either, but does happen. I don't remember the exact science, but I do remember the metaphor I came up with (evidence in my favor!).
Remembering something is like pulling a note out of a cabinet. It gets saved from the back, where the papers tend do degrade. Then, it gets put at the front. The trick is, the original note isn't put at the front. A copy is made. So remembering tends to reinforce, but also can change during this lossy copying process.
Jeez it was just a light-hearted comment. I think you can address your doubts about what I wrote with more kindness and less snark.
> None of them "churn".
By churn I mean "lose and hopefully be replaced". That definitely happens. Some brain cells die and are replaced, some live your whole life (or die first)[0].
The cabinet is just a simile for cued recall[1]. More recently learned/recalled items are remembered better (front of the cabinet)[2]. Recalling memories in a different context can change them[3], suggesting recall isn't a "read only" operation.