Well, I cranked the historical numbers. I made rough estimation of ratio of 2-pilots planes to 1-pilot planes, and subtracted it from monthly pilot numbers. 1941-1942 was pretty easy, but I did not want to search exact Navy:Army/Year production, so later estimations are pretty crude.
It is made on current-year production (when they actually beginning of training), not future year estimation. As a result, it is lower, as with war progression more fighters were produced, at the cost of heavier bombers.
The whole goal was to get the exact number of pilots, in exact year they appeared, so as game-engine works numbers may seems strange.
Mainly, the quantity jump was possible, because time of training was reduced. In 1941 (pre-war), pilots were getting 2 years of theoretical preparation (and 6 months of basic flying), after that 1 full year of flying.
In 1942 it was cut to 1.5 years of theory (probably also with basic flying), and 8-10 months of flying after that.
As numbers shows, maximum decline in-game experience for fresh pilots should be 5 points, as getting it bigger makes strange situation, where last months 10-12 class cadets were more experienced, than fully-trained fresh recruits.
In 1943 Army resigned from advanced training, and made special units to teach it after-graduation. This is the moment their experience falls. In September 1944 fuel shortages seriously hampered overall training, so from that moment both branches lose experience.
Here is modification. Branch, initial pools, and then experience, and monthly numbers. 1945/1946 numbers are simple halving, as there are no data about possible planes production, as war have ended before cadets could graduate.
ARMY. 800 40 165 35 155 30 750 30 350 25 175 25 90
NAVY 1350 40 130 35 165 35 620 35 310 30 155 25 75
I really hard searched for every possible pilot for Army, as initial calculations shown only 241 free pilots. It seems, that there was recently opened Pilot School in Manchukuo. Initial number of pilots was 30, and from 30 August 1940 there was regualr teaching. Hard to tell, what was number of cadets, but since there was mutiny of around 100 cadet-pilots in January 1941, conclusion shows, that during 4 months of 1940, on average, it should be at least 25 (25*4=100). So initial Army pool is increased by this number, but it was still not enough...
Luckily, there was also Manchukuo National Airways, paramilitary organization, which was mainly flying for IJArmy anyway (and have some volunteer-fighters), it should be possible to recruit them, as transport pilots in-case of emergency. Simple calculation shows, that organization had, at least, 43 1-pilot, and 24 2-pilots planes, not-used-by-Japan (as I do not have any date, when planes were acquired, so Japanese types could be delivered after war started).
Considering also, that there was Imperial Japanese Airways, mostly owned by government, and its civilian operation were suspended in January 1942, and there was also China Airways, they should easily provide 200-250 possible transport pilots, in case of emergency. So initial pool of 800 Army pilots seems reasonable (and if you are getting it below 200, you surely had to draw extra pilots into transport units, anyway).
So here are final numbers. You can also see, that actually IJArmy have NEVER more airmen, than IJNavy. It is also clear, that they were unable to train more, than IJNavy in any year. Hard to tell, what it says about pilot numbers (as BETTYs had really large crews), but even, when pilot numbers for IJArmy were greater, it could be only by small margin.