Why is the push to such high step rates (Grex and Arc100)?

I can't tell if this is turning into a peeing contest of specs or if
there is some actual merrit to these high rates. Assuming that these
are for machine tool motion control (yes I am well aware of other
types of motion control) I am failing to see the point.

With a possible frequency in the range of around 2 Mhz or more, this
works out to a pretty fantastic velocity. Assuming a resolution
of .0001"

2000000 * .0001 = 200 Inches per second
200 * 60 seconds = 12,000 Inches per minute

12,000 IPM is about 3 times faster than the fastest commercial CNC
machines on the market, and those use linear motors.

We know that stepper motors have a very finite, useful RPM range. The
steppers I have been toying with recently have a limit of 500 RPM,
beyond that torque drops very quickly. Assuming 200 steps per rev,
the maximum useful frequency is

200 X 500 RPM = 100000 / 60 = 1666 Steps per second

Even when factoring in microstepping of say 16 microsteps

1666 * 16 = 26,666 steps per second

Servo motors certianly can run at much higher RPM ranges - many around
4000 RPM. Putting that into some form of perspective relative to
motion, assuming a .2" pitch screw and a minimum resolution of .0001"

4000 RPM X .2" = 800 IPM (very fast for most industrial machine)

800 IPM / .0001" resolution = 8000000 steps per minute
8000000 / 60 seconds = 133,333 steps per second

So with the potential of stepping rates into the millions....

A stepper motor doesn't have enough torque at speed nor can it handle
that rate of stepping frequency.

A servo motor, driving a relatively fine pitch ball screw at a
resolution of .0001" doesn't need a stepping frequency in the millions.

Why is there such emphasis put on that level of performance in
stepping rates?

Chris (Confused as usual)