Put me in the MyWeigh GemPro 250 column. I have a D.I. MXX-123 ($300+). I recently purchased a MyWeigh GemPro 250 ($160).
I've now done extensive testing with both and if I had to give up one of them, it would be the D.I. MXX-123.
They're both accurate to .02 grains and both, in all my tests, measure EXACTLY the same whether it's weight sorting cases, bullets or powder. So why would I go with the GemPro?
1) Cost obviously (half what the AccuLab/Sartorius/Denver Instruments 123 costs)
2) Consistency
3) Workbench footprint (about half the width and length of the MXX-123)
4) Lifetime warranty
When I'm loading 100 rounds using the D.I., invariably, there will be 5 or 6 times that it does not go back to 0 after a dump. Usually this is small, fluctuating between .02-04, but sometimes as high as .08-.10. This results in a re-zero and do over on the last case. Doing 100 rounds is tedious enough without having a lot of do overs. With the GemPro, I've loaded 100 cases on two occasions and 50 cases on another. In those 250 rounds, I only had to re-zero on two occasions. In both those occasions, it settled on .02 rather than .00.
The other downside to the MXX-123 is how long it takes to stablize if it's ever completely unplugged. If you have one you've undoubtedly experienced this. You plug it in, turn it on and over about 30-45 minutes it will climb from its initial reading by up to 3 or 4 grains (not .00 grains or .0 grains but full grains). If you leave it plugged in, though turned off, the warm up time is significantly less, but still not zero. I turn my GemPro on and let it stand for 30 minutes, but this is just for my piece of mind. I don't think 30 minutes is actually required... maybe more like 5.
The only downside to the GemPro (which I've turned into a positive) is that it's re-calculation rate seems slower than the D.I. Say your initial dump is 43.00 grains and your target is 44.00 grains. If you trickle 5 kernels of Varget, it can be a couple of seconds before the digital readout reacts. What I do now is take my initial reading, trickle what I think I need to get to my target, then pick up the pan, shake it to even out the powder distribution and set it back on the GemPro. When I do this, the little "gr" flashes for a second then is steady. This usually gets me to within a kernel either way. I make that final add/subtract and dump... with confidence.
So, there's my input on the MXX-123 forum topic. Take it for what it's worth.