Update README with improved jump table dispatch detail

Clarified jump table dispatch description for better understanding.

Author: jgarzik

README.md

@@ -12,7 +12,7 @@ apython reads `.pyc` files and executes Python 3.12 bytecode directly — no CPy
 - **128-bit fat values** — inline integers, floats, bools and small strings in 16-byte (payload, tag) pairs, skipping heap allocation and refcounting entirely
 - **SmallStr optimization** — strings up to 15 bytes stored inline in the 128-bit value slot, zero allocation
 - **Raw Linux syscalls** — no libc dependency for I/O; buffered writes via direct `syscall`
-- **256-entry jump table dispatch** — single indirect jump per opcode
+- **256-entry jump table dispatch** — x86-BTB-friendly single indirect jump per opcode
 - **GMP for arbitrary precision** — big integers via libgmp when values exceed SmallInt range
 - **Reference counting + cycle-collecting GC** — deterministic memory management with a 3-generation collector for cycles
 - **Full async/await with io_uring** — high-speed async I/O via Linux io_uring (with epoll fallback), zero-copy TCP streams