I recently bought a 5V EAF to go with a new OTA. So I decided to measure it before mounting it.
I riggged up something to couple the EAF to a Rotary Shaft Encoder from Machine-DRO (UK company):

The encoder has a claimed accuracy of +/- 0.2 degrees. The coupler is a 6mm coupler from the EAF box. ASIcap (macOS) is used to drive the EAF (USB 3/USB C connector on a M1 MacBook Pro).
The 5V EAF apparently has 5760 EAF "steps" per 360º.
This is a plot of the offset from linearity as the EAF is rotated through a full rotation, in increments of 160 EAF steps (10º):

The abscissa is the nominal angle (i.e. units of 160 EAF steps) and the ordinate is the angle (in degrees) the EAF deviates from linearity.
There is no lost steps after 36*160 steps, the shaft encoder read precisely 0º again (although the encoder is practically friction free and I tried to align the coupler as closely as I could to being axially perfect).
There is about 0.88º peak-to-peak deviation from linearity across one rotation.
Next, is a measurement of backlash. For this, I stepped the EAF by 8 steps (nominally 0.5º) at a time until it has moved a total of 80 steps (5º); this point is shown in the abscissa as a "*".

Notice that after 80 EAF steps, the EAF has moved quite linearly from 0º to 5º of displacement. I then reversed the direction and no displacement is measured until the EAF has been asked to move by 32 EAF steps. If you look at the far right, the backlash appears to be about 3.75 units, or about 30 EAF steps (1.9 degrees).
For what its worth, the 5V EAF appears to have plenty of torque. Compared to my FSQ-85 and Feather Touch focuser, the new FOA-60Q focuser is very stiff. And the 5V EAF could drive it without any problem -- winter will tell if it keeps working well under low temperatures.
Chen