Actually, I disagree a bit with the tree concept containing multiples of the same limbs.
Let me explain my thoughts....
Going off of Wolff's concept of 2 or more 'mold pairs' (COO Intro here) and looking at pics of Joseph's Uzay molds (here
- and I'm not saying that is exactly how the Asian Kenner molds were), it makes more sense to me that if 2 molds were made per factory, then most likely one was in production while the other sat idle.
The first mold would run production of the whole figure for a set amount of figures. I say this because it makes sense to me that these molds would need periodic maintenance and inspection... lets say the mold should run and produce 10,000 figures before it needs to be removed and have XXX things checked, cleaned, lubed, etc.
While the first mold is removed for maintenance, the second mold goes into production. The second now would run for the same 10,000 figures. Rinse and repeat.
Seeing that many molds would be in use for many different figures, the maintenance department would have enough time to get it one and be ready by the time the other molds needs work and then swap it out.
The second mold would also serve as a backup in case of failure or issues.
So... if an order for 100,000 figures was placed, the idea of a 50/50 mix of the two is very possible.
But that is an ideal world. Factories might go longer or shorter between recommended intervals so numbers would be skewed from time to time, but in terms of us spotting a higher/lower percentage, we wouldn't notice much.
You could also have a factory that just said "run it until it breaks" and then we would have a much higher percentage of a single mold being used.
My thoughts on this aren't just the images of the Uzay mold or Wolff's guide.
It the things I have picked up on with my Leia Bespin endeavor and perhaps other variant collectors might have noticed as well is abnormalities in the mold.
For example: On Leia Bespin, COO family #1, turtle necks, I have seen many examples of a nick on the torso and in each time, the exact same COO stamp was there (i.e. 1x1A and never 1X1B).
If there was a true mix of all limbs and torsos, the nick should be spotted on 1X1Bs as well.
Now, I am sure that I might find one or do have one, but to me that would be a case of overstock and mix ups.
Anyway... I guess my whole point is rarity of a specific marking in a family can exist depending on the factory and how the mold was used and how many figures were actually cranked out using them.
I know my ideas might be skewed a little due to my Leia Bespin focus, but I recently finished my COO family/mold run and it came down to me finally finding a No COO Family #2 that was different from the other I kept finding (over 6 came through my hands and a ton of pics and listings seen).
I honestly believed for a while that perhaps that second mold was never used in production.
Anyway.... these are just my ideas and sorry if I rushed through this. I am more than happy to discuss this type of theory further since I love trying to figure this all out.