Scale that reads body fat

Not that accurate, though I've heard people say they can be helpful for tracking a trend.
Mine still thinks I have around 31% body fat at 5'7" and 130 lbs. I suppose it's possible but other methods (calipers, etc.) put me around 26-27%. It definitely went down when I lost 23 lbs though. I forget the exact number but it was in the high 30s 23 lbs ago.
My understanding is the way they work, they really only measure body half in the lower half of your body. I wonder if these can throw off the numbers depending on how you carry your weight. I think that might be a factor for me.

I would argue they are accurate enough.
What do you want them for is the question.

If you keep telling yourself, "its mostly muscle", they are accurate enough to say "no it b****y isnt"

I find they track your trend quite well if you use them correctly. It works by sending an electrical current through your body and it's measured based on the time it takes to get back to the scale. The current travels at a different speed through water. So, just like the scale weight you have to use it under the same body conditions or you'll get wildly different results. You'll need to track weight / BF% together and if you see weight change drastically one day (probably because of water weight) you'll probably see a wild change in the %.

I use an old Omron BFH-500 bathroom scale that produces readings both for %BF (body fat) and %MM (muscle mass); both readings fluctuate with hydration and with body weight. I've been recording those daily for more than five years. Recently, I decided to convert those percentages to pounds, based upon that day's weight in pounds; and, I discovered that, although they may fluctuate with that day's weight, the converted readings appear to be a fairly accurate trending tool--which is for what I had been using it, anyway.

Accurate? Lol no. Precise? IMO, yes.

By that, I mean: I have a scale that measures BF, water, muscle, and bone density via BIA. When I first stepped on and it gave me a percentage, I would in no way take that to the bank. But as I've stepped on other times, at 8 am after I've gone to the restroom before I've eaten and with all clothing removed, I've seen a general trends that would suggest a degree of precision. For instance, I've lost 14 pounds and started incorporating strength training. I noticed as I lost weight, BF went down. So did MM until I realized my protein was too low and that I should incorporate some strength training. Since, I've seen a .3% increase in MM along with leaner-looking progress pics and a lower BF reading. Likewise, I notice that when it's around "my time" and my weight goes up, my TBW reading also goes up. A few days later, TBW drops and I lose several pounds. The numbers are probably several percent off, but they are consistently that several percent off, which is all I need. I use it as a general tool to gauge my progress, not as a for sure, written in stone measure.

I will say that a LOT of the reviews I saw on Amazon for every BIA scale mentioned that those who are more athletic or who are lower in body fat will get crazy wild readings, even on the athlete mode.

Mine has varied as much as a couple of percent from one day to the next, which seems pretty unlikely to be accurate. The general direction/magnitude of the trend seems to track with my weight, from "stand back and squint at it" perspective.

I have a tanita one that is pretty accurate. The only accurate way to get it done is in the water tank or a pod. Calipers, for the most part, are not accurate. It's just a person using a caliper on chosen points of your body. When I used to get mine done in the water every 12 weeks, it was pretty much in line with my Tanita one if it was on the athlete setting.

I find my < $30 Health-O-Meter model BFM081DQ-63 scales readout BMI, % of Fat, % of hydration daily calories to eat to be relative accurate for the past 5 years.

walmart.com/ip/Healthometer-Glass-Body-Fat-Bath-Scale-with-Daily-Caloric-Intake-Athlete-Mode-BMI-and-Hydration-Level/14706457