Of course if you use analog scales, you eliminate the imprecision of digital rounding. Similarly with calipers. With dial calipers you can estimate well inside the 0.000" graduations.
Problem with the analog scales, depending on the angle, direction and lighting it could be less accurate then the digital rounding. The key I believe is consistence in every reading.
Still lots of sources of error in a mechanical scale. The pivot point for instance. Its type of construction jewel, kinfe blade, magnetic dampened lots of different types all with pros and cons and different sources of error. Nothing is error free, its a matter of listing the sources determining the type, and then working out an error map and uncertainty statement for that method of measurement.
For any reloading scales there are simple tests which check all sources of inaccuracy. If the primary purpose of the weighing is shot to shot powder charge consistency then just remove the pan full of powder from the scale, replace and reread a few times. If its also important that the absolute powder charge is accurate, then a metallic check weight very close to the charge weight is a good idea. If the scales are suspected of drifting during a session, (most often associated with electronic scales), then the check weight test may be necessary at intervals throughout the session. I use a Redding #2 scale, set up at eye level on a level surface and illuminated by a LED lamp. In reasonably sheltered air, it gives pretty much +/- 1 kernel consistency.
I have found my 10-10 scales to register a single kernel of RL-15 which is a pretty fine powder.
While I would love an electronic scale and auto throw for the purposes of speeding up the process I do not think my powder weighing method is holding me back as far as accuracy goes. My limited budget keeps going towards barrels and projectiles