It's been many years since I did 3D machining so take any advice I give you with a grain of salt and the advice of others. I'm not sure why you would need to have draft on a stamping die but I'll concede that you know the job better than I would. So much for my disclaimer:
When milling a draft the particulars (strengths/limitations) of the machine, set up, workpiece, and tooling all had to be taken into account. I always try to avoid tool deflection or induced vibration, especially when using carbide. A "Z" cut was used when the draft face was longer than the cutter flute length and a standard indexable insert cutter was used to waste the excess material. Hanging the quill (Kuraki boring mill) out that far would have vibrated badly if had not loaded the quill axis by cutting in the "Z" axis. I don't recall the SO but for the 1/2 inserts I believe it was 1/8 per pass. The biggest obstacle will be removal of excess material, the finish pass we used was always around .007 or so. I keep mentioning the vibration because if you're using carbide you'll need to use a rigid process to avoid fracturing at the cutting edge. The 1/2 inch circular inserts we used could be rotated a few degrees for a new edge when called for. I had thought that such a large radius would produce a poor cut but I used them several times to plow through material with better results than an insert with a smaller radius. We used HSS cutters with the draft angle freshly ground on for every job, a previously used draft cutter was asking for trouble. Speed was set by using the largest working diameter of the cutter, climb cut, and adjusted at the machine per the finish. The drafted wall was always cut in one pass where possible, the length of the cutter determines if this is possible. Where not possible, blending multiple cuts seldom produced a continuous wall surface and hand blending after was required. How critical your surfaces are will determine your requirements. Pocket corners with a drafted surface were always a call for adjusting the speed/feed as you approached the corner itself, watch for chatter. To me, hack out as much material as possible so you leave a minimal consistant amount the finish cutter has to take out. You don't want tool deflection because one region of the cutter is working harder than the other. If you can't get the entire wall in one cut and blending is required then leave about .0015-.003 for the blend-in.
This is what I was doing for large mold core/cavity work back in 1997 so if someone has more recent advice you can disregard what I've written. If you knew all this already then please excuse me covering what you already know. I hope this was of some help. Good luck.