So I wrote a program to simulate one million games of WiF and tabulate the occurrences of each weather type in each weather zone for each of the month pairs: S/O, N/D, J/F, etc.
This also makes it possible to measure the average number of impulses for each turn along with the chances for an odd number of total impulses, which will result in the side moving first having their Initiative chances for the next turn worsened. Note the numbers to the right of "Impulses" is the average total number to be expected - not the number each side gets.
One thing the results show is that the first S/O (S/O '39) is unlike all the rest. This is because it is the only turn with predefined weather for the first impulse. The results were so different that they appear in their own column below. Another strange thing is that although the average number of impulses in J/F is 5.23, it actually has the lowest probability of there being an odd total number of impulses. This means there must be many times where there are either 4 or 6 impulses in that turn. OTOH the chances for an odd number of impulses are pretty close to 50/50 throughout.
As for Japan, the results indicate that if you want to attack the NEI first (as many Japanese players do), and then later go for the Philippines, you should attack in a S/O turn.
