I did a run last evening without paying too much attention to details, but some things are maybe worth mentioning.
* In the test Toku's capital indeed had 4 archers defending it. However, in the test save it didn't have walls nor did we have to attack crossriver, so could take it with 8 axes. Can't remember whether it was on a hill but I think it wasn't.
I think we should indeed move all axes towards his capital immediately after DoW. Quite possibly we'll need them all. Any axes that lag behind might take on Osaka.
* Osaka got engulfed in our culture quite soon. It's good to build an Oracle in the AI's face.
Indeed it makes whipping more difficult for the AI and we may even flip it. I'm fairly confident that at least we'll get hold of the corn, even if Osaka produces some culture (from monument or religion).
* Road optimization silliness: I think it's most beneficial to hace a road 1N of GoldenFleece, 1E and 2E1S (corn). Almost all axes coming from the west with their 2 movement on roads end their turn at some point in GF. Then it takes a full turn to cross the river. If instead of ending their turn in GF they end 1N, they already have crossed the river. We can profit from the increased speed by having a road 1N of GF and further as described. This is especially of importance once the attack is underway our imminent.
* I'm wondering about those warriors we upgrade (if we do). Of course all warriors we have are without promotions but of course it would be cool to have a CR promo on them. Then the current warriors can remain as city garrisons/scouts/fogbusters. I guess it's also inherently more value for money to spend 80 gold on upgrading a warrior with XP than one without. So maybe in one of our barracks cities we can build 2 warriors at some point?
To take it even a bit further, could we leverage this by building barracks first in marble? Now in marble we'd be building promo-less axes, but rax+2 warriors is 90 hammers where 2 axes are 70, not that much of a difference. Besides any subsequent axes would also benefit from a promo of course. Arguably 3 promoed axes are about equal to 4 promoless ones.
* I think when cap is back to size 4 after settler whip and subsequent worker, it should gift the FP to marble and take bronze. Then when it's size 5 it can also gift corn if we build a farm somewhere for cap. Cap is then at max hammer configuration, first for axes then for mids, while marble can profit from rapid growth for whipping and/or possibly for hiring caste scientists not too much later.
* As I marched on Kyoto I switched GoldenFleece to 3 scientist with caste. However, by that point it has had the Oracle for 10+ turns and thus the chances for a GPr are rather high, despite the 3 scientists. In fact, I popped one. Nice for a theo bulb, but as I tried to argue the other day maybe not totally ideal for bulbing asap to engineering. On the other hand, it can't really be avoided, so maybe we should just leave that up to the RNG.
If we get a priest, possibly we just have to self tech MC. However we had to self tech theo in the other scenario so that's maybe not too much of a difference.