Conclusions

Static resistance is significant - about 20 lb. This is roughly consistent with gantry weight of 60-80 lb and 0.1-0.3 friction coefficient of oiled bronze bushings.
Thus, motor strength is important and for 300 oz in steppers - geardowns is suggested.
Double sided drive may help - not tested yet. It will help for tooth jump force - it's a way to increase drive other than via 2x geardown and is a simple solution in our current system to get higher force at $40 per axis, with minimum addition of system complexity.
Steel belts are definitely helpful. Nylon belts - before motion - belt stretch happens and backlash is on the order of 1-5 mm on 6' 3D printer. With steel belt - backlash is 1-2 mm
We can improve system performance by decreasing resistance. V-groove bearings would achieve this.
For higher machining force in general - we need geardown, stronger motors, more motors - in addition to the less resitance.
Practicality limits are double sided drive and dual axis per degree of freedom - 4 motors total. For machines of larger force - geardown would be essential.

Blade Plots

11/14/21 gets us 13-23 lb friction on Y1. At times static friction is overwhelming and axis doesn't start. X motion is stronger. 20 lb stopping force is available at 500 mm/sec on x and y. Z had no problems (due to slow speed?)
Y1 has more friction than Y2 - due to autoparallel-Y piece being on Y1 side and less stiff? Y1 axis rotates around the Y axis as does slide into y-autoparallel piece. Y2 doesn't due to stiff x connection.
Without increasing force, we are limited to safe zone of 2000 mm/sec?
- Data collection: measure upper speed where drive force remains 20 lb throughout entire length of axis - on x, y, and z. This is 30% of theoretical Max of about 64 lb drive (300 is in)

Y1 belt kept rubbing on idler side. Idler was slightly slanted, and wobbly pulley on x motor did not help.
Enlarging hole in pulley from 6 to 8 mm is difficult with mag drill. Did it in 0.4 mm drill increments - drilling is not self-alogning even with this small increment
Z pulley wobbly and bent from too string vice hold
Z and y1 pulleys replaced - these were wobbly
Drilling was done in 1/64" (0.4 mm) increments starting with 1/4". Motor shafts are 8 mm, pulleys were 6 mm
X autoparallel is implemented with oversized x mount holes on x1
Y autoparallel is implemented with autoparallel y 3D prints and nylon bushings; nylon bushings are longer than needed
Y and Z only use hollow shaft. Hollow shaft uses nylon bushings, solid shaft uses bronze oiled bushings

At half inch cutting speeds of 20 ipm or 500 mm/min - this is prior to pulley size calibration of about 30% larger pulley
11/12/21 gets us 25 lb towards negative y1, 20+ for both directions of y2, and 20+ lb for both directions of x
Positive y1 - data?

Use larger motors? 4.2A 425 is in [2]. This is 71 lb holding with 3/4" pulleys
Low current version [3]. 54 lb. Larger motor. Has almost 20 lb extra drive. Strategy - tooth jump makes HTD 3 mm pitch not a great option. Double sided drive would be a better solution. However, at higher speed (3000 mm/sec), torque reduction of motor is significant. Is that the drawback of low-current motors? More rapid drop of torque-rpm curve?
This article implies that we may need higher voltage in case of rapid dropoff of torque - [4]