Musings on screw pitches

All this talk of fast machines, and some recent events, have got me thinking
about stepper based systems.

Maybe it's from reading the web published stuff, or maybe it's from looking
at pictures of other peoples hardware, or maybe it's because I've been
confusing torque and power, but I have long had it in my head that a 4 or
5TPI screw was "about right" for an NEMA 23 double stack motor driving a
small machine.
Certainly there are a lot of machines with ballscrews of about that pitch
(or finer) out there.

A buddy recently built a small router (envelope about 400mm x 400mm x 100mm)
the screws he had were about 20mm pitch. I speculated early on that it would
go "a bit" faster than my, very similar, machines 2500mm/min. And that this
is because the motors would have more torque at the lower rotational speed.
Nb that's a metric "bit".

Well it went 6 times faster. http://www.caswa.com/nobott_on_speed.wmv
(cutting the road runner)
Not only that, it accelerated 6 times faster. Same drives, same motors.

The acceleration is the clincher. It's clear that, on machines of this size,
that the mechanical advantage of the ballscrew reduces the mass that the
motor "sees" to a point well bellow the inertia of the motor+ screw.

When I compare the performance of my friends new machine to mine in terms of
RPM an RPM/sec, they are pretty close.
OK, the rest of the discussion (resolution, microstepping etc) pans out from
here.

My question is, why is it not in the FAQ that you should use 20mm pitch
screws on small machines (particularly routers, where 0.05mm for a half step
is plenty of resolution)??
Opinions..

PK