^^ I have found with mine ($70 one) that if I put an initial weight on, like a folder square of cardboard, it will weigh out to within +-1mg every time. Usually it starts to drift and can get away, but the initial reading it always the same or 1 higher or lower. If you don't use tare, adding to the cardboard is really quite accurate. I just make about a dozen measurements and note the weight at the beginning, and I get a result of +-1mg. For things like DOX, I weigh out a large mass and put it in liquid, but for things like 2C-B, AMT, 4-AcO-DMT, and so forth, it works great and I've never had any accidents with mis-weighing doses. If you just put the chemicals on the scale when it reads 0, then it's not going to work well. So you work around that. I mean, even if the scale is wrong, and the cardboard reads 330mg when it's really 350mg, you're just measuring the difference in weight between the plain cardboard and the cardboard with crystals on it. Let's say that the scale is 20mg off with the carcboard as I mentioned, which means that each mg it measures is roughly 330/350 = .94mg. Then measuring out a 20mg dose may be off 1mg, reading 349 instead of 370. But that gives a NET weight of the chemical at 19mg, which is a small margin or error. If it's a chemical where that's an unacceptable margin of error, then use liquid measurement. But for most things, the difference between 19 and 20mg is not significant.
I mean, yeah, everyone should buy an expensive scale, really. But not everyone has that kind of money, and it's tremendously better to buy a cheaper one than to not have one at all. No matter what anyone says, I stand by my scale, the JPrecision gemological gem scale 10gx.001g.
But I do agree that scales which only measure in 10mg or greater increments, or really even ones that measure in 2mg increments, are too ineffective. As well as some brands of 1mg scales.
Sometimes harm reduction is not perfect, and it's best to talk about plan B. If most people are never going to drop hundreds of dollars on a scale, then we should accept that no matter what it's going to happen, and begin discussing $70 scales that are good enough for most purposes if used carefully.