That code has a potential flaw in it. I'd like to see what it does, but I suspect I'll need to make a change.
Whenever it sees the index pulse it will zero the step count out. In theory, deceleration will mean that we will see index at the same place on the gauge every time at exactly 25600 steps since we last saw it.
Today's testing does not bear that out. We know though, that we can step for hours by counting steps with complete confidence in our accuracy. I anticipate that I will need to use the index pulse to get an initial position and ignore it forever after, as we discussed. We will need to establish a homing speed that is repeatable.
v4 likely won't give us repeatability, but it'll be interesting to see how much better it is. I'll bet a few dollars of your money that I'll be making v5 shortly after you start testing tomorrow 