Okay, I did some research and I found out the production numbers in my head which I used as a reference were from late 1943... 120 is indeed way to optimistic for Chosen. So lets assume 30-40 of all types for now.
*Nods*.
I'll explain a bit of my understanding about tank production. (This is intended for information-sharing, not a diatribe.) The way I tend to look at tank production for my countries is to figure out a work-flow. As I'm now currently employed in a small industrial setting, I am seeing a lot more of how these sorts of things work. A tank requires about thirty thousand individual parts, usually manufactured by between a thousand to two thousand different companies or factories - give or take. And the factories are pretty specialized, requiring very heavy cranes: twenty to thirty tons capacity - that's a lot bigger than anything my factory has, and we manufacture
massive industrial equipment. When my company builds a lot of something - for instance, we build containers for the military to ship parts in - we can get a pretty good workflow going. The welders get to know the product; errors the engineering drawings get fixed; and purchasing gets parts delivered regularly so that the stock stays up to date. Economies of scale. When we do something custom - and we do a lot of custom work - the process drags. An engineering drawing ends up wrong but isn't discovered until the pressmen make it - throw it away and start over. The welders have to go figure out what they're supposed to do. Everything bogs down.
With so many parts, including highly-complex members (engine, transmission, steering, suspension, turret...), workflow becomes important. The factory thus has to establish a goal: let's say they promise one tank a day. It takes eight thousand man-hours to build this tank, and the factory has a thousand shop-workers (by which I mean people actively engaged in assembly). This means that a thousand skilled workers can assemble one tank per day. (In real life, these thousand workers would be likely split between teams working in a single production bay, each building many tanks over the course of days or weeks - but you still end up with one finished tank completing per day.) What I know from this environment implies to me that new workers will slow production down up to three months after they're brought onto the manufacturing force.
When I set up my estimates for tank production in France, I pulled a list of the number of working days in each month, and then established a rate of production for a factory. Here's an example:
January (1943): 21 days
February (1943): 20 days
March (1943): 23 days
April (1943): 22 days
May (1943): 21 days
June (1943): 22 days
July (1943): 22 days
August (1943): 22 days
September (1943): 22 days
October (1943): 21 days
November (1943): 22 days
December (1943): 23 days
If our hypothetical factory produces one tank per workday, then it produces 261 tanks over the course of a year. This seems to be about the rate where tank production becomes efficient, and so I tend to aim at that for my baseline. In France, I prefer to manufacture at a slightly higher rate - 1.5x to 2x tanks per workday - to keep one or two factories running for domestic and export consumption at peacetime rates. (It's my estimation that, in wartime, France should be able to build, give or take, between ten to fourteen tanks per day from the combination of the Renault, AMX, Somua, ACL, and ARL factories. This is just bringing the five factories on-line and up to standard work capacity. In true desperation, the French could probably shift light tank production to truck and railway rolling stock manufacturers in order to free up a bit more capacity. But I don't believe that will ever happen in Wesworld, short of a Great Power going nutters.)