I have used the RCBS and the Hornady extensively. Although I prefer the Hornady, neither one is particularly awesome. Both claim to be accurate to 0.1gr, but neither one is anywhere near that. In fact, no digital scsle I have ever used are reliable. Nothing beats an old fashioned balance beam.
The problem is that the scale reads to 0.1, but the internal electronics are not up to the task. Even though it might say 0.1, it could be as much as 0.3 or 0.4 away from that. I have tried several models of each type comparing against several balance beams and a laboratory grade scale accurate to 0.01 and never found one I like. The references all agree, but the digital scales and powder measures don't. And yes, I have been careful with level, and allowed lots of time for temperature stabilization.
That said, I almost always use my Hornady even when doing load development. I slow the speed down, increase the trickle timing, and set it to throw charges a few tenths lower than I need, then throw that into my favorite balance beam measure and use a manual powder trickler to bring it up to zero. Slower than using a measure, faster than using a scale, but way more accurate and not all that onerous.
Another observation. How well either measure works is extremely dependant on the powder you are measuring. Big grained powders don't work nearly as well as the fine stuff.
Last comment - DON'T LET POWDER GET DOWN IN THE MECHANISM...... Don't ask how I know. .....