I agree that any sizes will be random. I have posted my numbers way back June of this year to show how frequently I get sizes other than "Normal" sizes by weekly. I did not include the number of "Normal" sizes. This time, I included it, but did not do it by weekly, but by the whole population.
One thing I can say, I can get my numbers accurately because I'm a hoarder of all my boxes so I can give a visual presentation of my cattery population. I don't put my sizes for sale at all so here are my numbers. Please remember, this is my population, my sample size, not yours. I am only giving a visual presentation how frequently I get sizes. As you can see below, I have a total of 2269 boxes. I did not include any collector kitties I bought. This is strictly what I got from pairing my kitties.
As you can see above, I have more chances of getting more "Teacup" sizes than "Megapuss" sizes. The percentage between "Bigger de Big", "Petite", "Toy", and "Teacup" are quite close. There are no differences.
The next image I am going to present one of my kitties, Reynard, which is a megapuss. Many assume that you could possibly get different sizes if you breed other sizes to another sizes and get non "Normal" sizes for the babies.
With Reynard, I was lucky enough to get 2 "Megapuss" sizes and 1 "Teacup". I never birthed these kitties so to do a proof of concept whether sizes pass to the babies or not, I cannot tell you. For one, I will not give a biased experiment. If I really want to do a proof of concept (POC) and give an accurate and exact sample size and detail numbers, which includes percentage, I will do so. I live and breathe statistics as it is part of my real life job and I am a programmer as well. I love numbers and data.
Hope the visual presentation helps you. Have a great day, everyone.