the reason is in incorrect input.
Undoubtedly, the question is simply what variable is SO wrong?
I have done my share of lobbing 308 and 6.5mm to "impossible" range, almost always "sighting in" on the spot, as temperature measuerement is by nose and ass, and pressure by weater forecast, etc.. the 6.5 becomes effectively unspottable beyond 1500m as we do not have desert sand here.
I have been pestering Yrcan over this and i am convinced that all the normal variables are accounted for and used well within accuracy that should yield less than 1/10 of the error we are seeing.
Running my ovn dope i get between 40.0 and 44.0 mil UP, considering the probable met/env at Yrcans site.
Temperature: 1°C is 0.015mil /
Pressure: 1hPa is 0.044mil
Direction: 90 vs 270° is 0.5mil
headwind: 1m/s is 0.06mil
velocity: 1m/s is 0.11mil
Humidity: miniscule effect, and anyways close to 100% in all Yrcans shooting
BC: 0.010 of G7 BC is 1.2mil (in comparison it is 0.3@1350m)
Range: 10m is abouts 0.44mil
As far as i can tell from experience with measuring all these variables, at distances beyond handheld lasers, range is by far the most challenging. I estimate Yrcans temp, press. and other measuerements to be accurate within 2-3units of measurement. Wind probably far better, as we can see there is very little wind anyway.
Range has been remeasured using different maps, GPS, etc.
For any probable error in variables other than BC, the range must be over 2300m, which it is not.
Thus, as i see it we are left with BC, obviously very important at this flight time (5.3s)
To add another issue to the equation is that there are variables that affect known variables, changing what you thought to be correct, especially at extreme range. Things like BC changes over the trajectory
The question becomes then, does the G7 BC change with velocity, and does it change on the order of 0.02 units?
Does it change at the speed of sound? After all the bullet is travelling 600m, and for 2 seconds, below the speed of sound. If this is the case, the only solution will be a custom drag model, or to tweak it with different BCs for different ranges, or to use a bullet that fits the existing drag model.
Maybe the bullet enters a region of dynamic instability, and is coning , or some kind of limit cycle oscillatons - thus increasing drag?