diff --git a/Roadmap-for-Flexibly-Timed-Platforms.md b/Roadmap-for-Flexibly-Timed-Platforms.md
index e4a2bc9..fa8a028 100644
--- a/Roadmap-for-Flexibly-Timed-Platforms.md
+++ b/Roadmap-for-Flexibly-Timed-Platforms.md
@@ -55,8 +55,11 @@ With a workable pattern suitably discovered, apply the same to other architectur
 
 Both the Z80 and 6502 both almost have instruction decoders themselves, in the form of the disassemblers. It would probably make sense to formalise those; it also wouldn't necessarily be unproductive:
 
-* in 6502 world, it would simplify the logic flow for instruction set selection between the 6502, various 65C02s and possibly the 65816; and
-* for the purposes of the Z80, it might aid in unification of CLK and CP/M for OS X.
+* Apple used embedded 6502 cores for various helper chips — the IOPs in the Quadras, the ADB controller in the IIgs and elsewhere, the latter of which I was seriously considering a static recompilation of in any event, so that could be an alternative baby step into JIT threaded world;
+* also in 6502 world, it would simplify the logic flow for instruction set selection between the 6502, various 65C02s and possibly the 65816; and
+* for the purposes of the Z80, it might aid in unification of CLK and CP/M for OS X (and, from there, maybe adding support for CP/M-86 as a route into the 8086?).
+
+If a JITty implementation of the IIgs's ADB controller goes well, it might also be the smart move to add an HD6301 for the benefit of the ST and a better IKBD emulation.
 
 # Practicality of Templating