I've only had a quick look at this thread and the circuit, but imo the pic is simply controlling VRef. It's possible it is modulating it to provide variable VRef - but I think this unlikely since there's not a lot of point with the voltage divider if that's the case.

What I've done in the past to control a 297 is run the output of the PWM through a two stage RC filter to set the current, but doing this means there is no need for anything else around the VRef pin.

I think if I were working on that circuit I'd consider replacing the 297 with the PIC (albeit a slightly fatter pic). I'd use one of the comparators along with the PWM brought out as above (other side fed via the the current sensing resistor as per normal). Think I'd probably do most of the work in the interrupt probably using change on port ints to detect comparator and clock change. Should be possible to include microstepping etc - along with the current control stuff.

Failing that, why try and match exactly the circuit? Do as others above have suggested and simply use a 555 as a charge pump to keep the motor in high power mode.
All you really have is two supplies to your VRef pin - one permanently at 5v the other switching 0 to 5v. Run them both through a resistor (possibly the 555 output through a diode to make it easier to figure) and to the potential divider. If the resistor values are matched then roughly speaking when both are on the current will be double that when only one is on. Obviously you can chose resister values that give you the ratio's you're after.

Hope that makes sense.


[edit] I see above there's a circuit using the 555 ;-)