Flashbacks to when I implemented a new storage system for a Lisp/Prolog system while porting the core code from Pascal to C. We stuck with linked list of pages so we could keep pointers and gain some speed over array indexing for every object access.
No doubt. The second attempt is always better. My initial plan was to write a complete first implementation in C so that it's always possible to bootstrap the language, then write a compiler inside the lisp itself, or write a Rust version. Hope I somehow manage to do it all before I die of old age.
Aha I've been semi-vibe-coding a scheme for esp32c3 and linux at the same time, and forgoing the libc too, so baremetal c. I went with a slightly awkward approach for allocation, with heap being seen as pages, and within pages, fixed size objects of size 8, 16 or 32 bytes. A pair is two words, and either bitpacked or pointer to another object. I'm not far in, vaguely following the Peter Michaux approach.
Um, see The Second System Effect.
https://en.wikipedia.org/wiki/Second-system_effect