I've got a little experience with Matlab and was bored so I decided to test this out. My methods below are from an intro college course of Matlab and it may not be efficient and it may be wrong. I'll do my best to explain what's going on. First I'm going to explain the significance of the numbers used:

1,000,000: Group of the population. Not much significance.

3500: Take the group of the population and multiply it by 3500 to get roughly the world population of females able to reproduce. I did this since Matlab sadly cannot handle doing a random number generator for 3.5 billion numbers(haha).

75: I pulled up a random page that says 66.57 is the average death age of the world today so I took 5000 devided it by 66.57 and got 75, the number of generations to test.

Basically what I did(or tried to do) was take a segment of the population and generate a random number between 0 and 1. If the number was less than .5 it would round down to 0, if it was equal to or greater it would goto 1. It would then repeat this for 75 total generations. It would then search all these people to find whoever managed to have a boy every time, signified by the number 1, and add it into the total counter. This would then repeat for the entire population.

Code:

for Y = 1:3500
POP=round(rand(1000,1000));
X=0;
while(X<75)
X=X+1;
sum(sum(POP));
for A = 1:1000000
if POP(A)==1
POP(A)=round(rand(POP(A)));
end
end
end
total = sum(sum(POP))
end

PS, it takes about 5 seconds on my computer, with 4GB of ram to go through all the way once which means it would take about 5 hours for me to display a result (5 seconds * 3500 tests / 60 seconds / 60 minutes). I'm totally not doing this but anyone that wants to maybe clean up the code or let it run I'd like to see the result.