It's an interesting questing. Maybe there are too many factors for them to supply an accurate figure?

If you are chasing accuracy of the 'servo' isn't that effectively determined by it's resolution which would be a function of motor steps per revolution, gear ratio, pot accuracy, motor driver circuit output frequency)? Versus the fixed resolution of the signals sent from the Radio (whether it's 1024 0r 2048 bit) or the FBL controllers output.

Modern digital servos are generally high accuracy (high res) hence the move towards direct to swash applications where as in days gone buy a great way to increase the resolution on any given control was by using appropriately designed bell crank systems.

 

so maybe servos are so high rex that inaccuracies come from somewhere else? I thought while using cheaper servos that i could identify the position was not allays the same from one pitch change to the next. Ie the accuracy from one cycle to another was not accurate - ie did not feel in the same position. Also the swash was varying in level through the range. So I dont know weather this is resolution or inconsistency in the build. Its really interesting how some manufactures quote high resolution servos but dont actually quote anything to measure this attribute... http://hitecrcd.com/products/servos/ultra-premium-high-resolution-digital-servos

 

A cheap servo will be more likely to have slop in gear mesh, gear shafts, pot mounting, etc. A minute amount of slack in the pot and related coupling to the output shaft will cause inconsistent indexing of servo, and center point (neutral) will fluctuate.