Peter, you are absolutely right with gearboxes being required for proper load matching in most cases. My point though, is that when discussing a DC Motor, or an AC Motor, it seems that all of my customers will choose a standard motor, and any old gearbox to give 'sorta close' to the desired speed output.
On the other hand, in cases where the end user is mechanically sharp, or has real reasons for wanting a servo system, they will make sure that the motor and gear train and load are all well matched for each other.
Heck, I have one customer that uses the same 4 speed gearbox on just about every piece of equipment they own. They must have thousands of them. If they can't get the required output speed, they'll stick two in series, or change drive pulley ratio's, but they never ever ever EVER give any thought to actually matching prime mover load and WK2 to the actual load. <shudder>
Just looking at speed loop tuning parameters, you can get a pretty good idea on how well motor/load are matched.... really low gains generally indicates pretty poor mechanical match. In those cases, you can certainly stick a servo motor/drive on, but will get the same horrible performance.
I don't know, perhaps people are just getting lazier <grin>. Too often lately it seems that the folk doing the electrical controls are required to go to great lengths to 'tune out' mechanical problems, because, obviously, it's easier to change a number then bolts and gears. /sigh
(Edit)
Oh, and the simplest way to think about the benefits of inertia matching is to think of simple electrical circuits. Consider if you will, a battery and a resistor. You get the most stable, and maximal power transfer from source to load when the resistor value equals the internal source resistance of the battery.
In other words, when load impedance = source impedance. In most cases, mechanical inertia can be thought of directly as electrical impedance.