Thought of two ways of testing this:
1. Look at the market situation of a single product. Divide the shortage (surplus) by the amount a single corp produces in a month. This will tell you how many corps are necessary to be built (closed) to close the gap.
Now, multiply the Corps in Deficiency (Excess) by the number of workers needed to supply a single factory of the product in question. With this you will know how many workers need to be relocated to the market.
Repeat the step for all products.
Now, what do you do? Sum of all the numbers. If the number is 0, then yes you can have a closed economy. If it is negative, then you would need an infinite amount of pop to be in autarky. If positive, then you need 0 population.
Of course, there are disturbances in the short-run or some aspect unique to the period in question, so it is highly improbably that it will be exactly 0, even if you could have autarky. So, what to do? i) leave weapons out of the analysis, ii) use an average over several periods, iii) use hypothesis testing.
2. Repeat the step for all worlds.
Order the numbers in ascending order. Order the worlds in ascending order by population. These should match.
If the order does not match, it is a possible result of the disturbances or you may indeed have a (slim) chance at autarky.
I suggest someone runs the numbers before adding to this conversation. If many of you are that interested in this, then you can use Google Drive to coordinate and gather the data simultaneously.