A few weeks back I was putting together some load ladders for load development........using the dinky cheap Lee Perfect Powder Measure (PPM) to get close, would then trickle up with weight measured on a beam scale, which was so accurate that only a couple kernels of H4831sc would bump the scale 0.1 grains at 47 grain level. The Lee PPM consistently dropped within that 0.1 grain level, and if off, was usually the 0.1 grain under. There is a guy on the High Road Forum that has run tests and weighed over 50 drops and found the same thing. They are that accurate. You would not think so looking at them, but they are. And the Lee PPM Deluxe is very slightly better.
Now comes the fun part.........load increments used in those load ladders was 0.3 grains. When you get the load ladder right, overlap of velocity in those nodes is in range of 0.5 grains. So being off 0.1 grain is irrelevant. Calculate what the velocities are from min to max..........divide charges to get FPS per 0.1 grain, the run that thru a ballistics calculator and you find practically zero difference in elevation of bullet impact out to whatever practical distance you might be shooting to.
Bottom line is the LEE PPM is what I'm using......and it works for what I'm doing. There is a knack to use it, however, and that mostly relates to slowing down the process, so as to give powder time to fill the cylinder.......and time for it to dump. Instead of a hurried up/down......it's an up.....pause.......then down.
On the digital setups, I'd be curious to see what results are if you take 20 drops from one of those, then put that powder charge on a beam scale. Will that be dead nuts on.......or does that also vary by 0.1 grains.......or more.