Hello Henk,
I did not yet understand all you posted.
Accelleration Problems I had too sometimes. I just "fixed" that in the frequency inverter for the Spindle. (slowlyslowly ramping up, maximum speed of the machine is now 600 rpm with a much smaller motor and 1:5 gearbox ). The bigger motor was better for milling, but had no torque at lower speeds.
For the gear blank rotaty axis I just switch the spindle on and then wait until the spindle runs for a minute or so. "ugly fix, but the stepper on the rotary axis just cannot keep up at certains speeds"
The hal file I more or less, bluntly copied from Andy Pugh

There is one scale that connects the gear blank axis to the x axis
( they are prallel on my machine). I adjusted everything so that if the value of axis.1.scale is 1 it will rotate the gear blank a full circle for every mm moved by the x axis.
Then for a metric gear. The value for axis.1.scale is
axis.1.scale = tan(gear_angle) / (Pi()*number_of_tooth*module)/cos(gear_angle)) = tan(gear_angle) / Pi()*pcd
At the moment I just calculated that with a calculator and feed it into the hal configuration manually.
That means the axis.1.scale changes for every gear blank based on module, number of teeth and helix angle.
Then the gear blank rotates according to the x axis travel and the downward motion -z of the hob ( see my earlier picture) is not required anymore and the hob travels only prallel to the gear blank axis.
I tried this since last post by first cutting a set of gears based on the sketch from previous post. Then driving hob and gear based on the new method.
Now next step would be a proper hob shifting based on z-axis travel to wear the hob evenly.
Greetings Timo