Charge turbulence is the main governing factor for combustion speed, and charge motion goes up with revs, leading to combustion taking place over a certain number of crank degrees rather than a certain amount of time at higher revs, thus ignition timing in terms of crank degrees doesn't need to advance further.
however its weird that combustion speed should behave exactly like that (lagging and then exactly mirroring crank speed from a certain rpm up). besides most (many?) maps have a plateau (identical numbers) not only in the upper third of the rpm range but in the upper third of the density range also. that doesnt make sense.
except assuming lazyness on the part of the programmer, what is the reason?
Well, when timing is near optimum (for maximum thermal efficiency), the curve of thermal efficiency vs timing is fairly flat, so a degree or two away from optimum has little consequence. I wouldn't call it lazyness, more like spending your time where it counts the most
"Schiefgehen wird, was schiefgehen kann" - das Murphygesetz
Still takes roughly the same number of crank degrees for combustion whether the cylinder is half full or completely full - hence, comparable ignition timing. You can be certain that with any engine built in the last few decades, the designers and calibration departments have put some thought into this. Keep in mind that setting the ignition timing for best torque is not the only consideration for how it needs to be set.
The upper third of the load range also sees the mixture getting richer. On late cars going from 1.0 in all the light to medium load regions to 0.85-0.9 at full load. The extra fuel has a significant ant-knock benefit so timing doesn't need to be retarded as agressively.
in the meantime i realised that to some extent the answer is:
because lately the efficiency maps are being modeled as a function of mass (rather than of pressure) the effect of decreasing cylinder filling after the torque peak disappears.
. . and of course if mass is holding up while pressure drops, temperature must be coming down and we know the effect of charge temperature on knock and MBT.
