Interesting things you have about manpower and labor pool. It is true that historically Japan did not really 'mobilize' in the same sense as other countries. They were able to muster over 3 million troops on the home islands in their final mobilization to defend the home islands, and did not send as many troops as they could to the China theater as they can in game under the same mobilization system (supposedly according to sources I read due to shortages of weapons to send any additional troops outside of Japan).
It's also good to avoid having too many small flavor events that occur regardless of events in game. Ex. don't have something occur if the conditions for it doesn't account. I think having the player decide directly on who has political influence (IJA/IJN) is a bad idea as well.
Hoping for the best of this mod. If it turns out well, who knows, maybe it'll be included in 1.05. I intend to rework Sino-SovietSino-Japanese War for 1.05. I don't know enough about Japan from a Japanese perspective (given I can only read Chinese and English).
Side note: Interested to see that I can understand all of the Japanese as it is kanji. heh
Yes, modding is for your own personal enjoyment. I do it for that reason as well (I've stopped modding though, mostly only working on core). For the new OOB in 1.04, Matt and I spent a considerable amount of time on researching Chinese and Japanese OOB (had to look on Japanese websites and try to understand it via google translate) and adapting it to the way DH represents units (there are some mods that don't understand the way DH models units, I don't always either which was why it was advantageous for me to have the ability to ask the team about what a 'Infantry' division represents. Also each country's divisions can vary, even if the names for them are the same type of division, many of the choices were based on brigade or regiment composition instead of the division name). China does have a huge amount of militia, yes, but the new OOB is historic in numbers. China actually suffers from an IC problem in 1.04 due to most of their IC being spent on supplies. In addition, a good portion of the troops are 'locked' to represent the fact that those militia are the personal troops of local warlords and not under central authority. They get 'unlocked' after a period of time, so Japan has the ability to overrun China and destroy the large stacks at the beginning of the war (AI fails at this). Japan also gains control over Shanghai and can rush the Chinese capital (AI fails at this).
Regarding the mass production of militia afterwards, I do have ideas about that (none of this is implemented for 1.04). I wanted to introduce a risk effect to it, which would involve reworking the Chinese Civil War. Specifically, the more CHI mobilizes and drafts militia historically (forced conscription, no supplies, forced marches), the more 'post war dissent' they would get, so if CHI does win the war against Japan due to heavy mobilization, it results in a very weak position to fight the communists afterward (mobilization, the way it was carried out, was very unpopular, but due to war unity reasons, all parties agreed to support the KMT during the war). If allies provide weapons and training (ex. in India), then CHI gets some troops without post war dissent risk (but this is a cost for allies who have to donate supplies and IC, and this would take longer than emergency conscription, as historically the allies never even finished training the final batch of units in India on time before Japan surrendered). The ultimate idea was to balance it so that CHI would not be able to produce enough units by IC alone if facing against a good JAP player and have to use the emergency conscription system (implemented via event, as they did historically) to stay alive until the allies bail them out (via lend-lease). There would be no event that decreases militia production time. In fact, I'm even thinking of an event to increase production time to simulate lack of heavy industry in china to produce weapons. Japan should not be forced to mass produce Infantry because historically they did not send anywhere near half of their 'mobilization' potential to China. Instead, Japan should figure out how to strategically capture important regions in China to reach a negotiated peace (discouraging occupying territory via a partisan system where Japan can lose supplies (which is then gifted to CHI or CHC, and can be used to conscript troops via emergency conscription system) or face a mass uprising if it occupies too much territory and does not assign sufficient troops to pacification. Historically only ~10,000 troops were assigned to counter-insurgency vs an insurgency numbering 450,000 communist militia).
A peace system would also be implemented (ex. if too much conscription or all major cities occupied, China may capitulate) to prevent the ahistorical complete annexation of China that occurs in games (Japan simply wouldn't be able to police the area), along with the partisan system remaining after the war so that a puppet Chinese government would need to keep their troops maintaining order in China rather than conquering the USA for Japan. The peace system can even be tied to the control of IJA/IJN (ex. one of the earlier peace proposal process failed due to the IJA believing they could win in 3 months).
Anyway, all of this was just ideas, I don't even know if I would have the time to implement them, or refine them. Just giving you food for thought.
(Edit: Also many of these systems are aimed at players, there would need to be tweaks for AI if it can't handle certain aspects, like being less rough on post-war dissent for AI CHI if Japan is a human and CHC is human as well, or being less rough on AI Japan if CHI is human in partisan system).
(Edit 2: Fixed error, meant sino-japanese)
