Bodyfat % fix via SW update possible?

Seems like it should be fixable with software, especially since it seems like it's universally measuring waaaay too high for pretty much everyone.

I just picked up the S2 on friday, upgrading from an Index 1. I knew going into it that the BF% wouldn't be that accurate, but I guess I thought maybe it would be "close". Nope.

At 5'10", 162 pounds, the S2 thinks I have 21% body fat. The next day, when I'm down to 160 pounds, the S2 for some reason thinks I now have 22% body fat. 22% body fat is close to obese. Trust me, I'm not obese. And neither would ANYONE who is 5'10" and only weighs 160 pounds.

Something is just straight up wrong about their algorithm. And it wasn't this bad on Index 1, which was regularly showing my body fat around 14-15% up until I switched to the S2. In reality I'm probably 12-13% or so, so neither scale is correct, but the S2 is vastly, vastly more incorrect. It is incorrect to the point that the number it gives can't possibly be true for any human being with my height and weight.

I'm not expecting it to be accurate to within +/- 0.1% or anything like that, I get that it's not a medical-grade device. But to be wrong by 100% or more, for almost everyone, makes it an embarrassingly bad algorithm that should be able to be easily improved quickly by someone at Garmin. Not sure why it's taking them so long to even improve it a little. Seems like, by now, there should have been an update that makes it work at least a little better.

It should indeed be fixable via SW update. The bodyfat issue is pretty much the only reason why i haven't purchased this scale yet.

It's possible, I guess, that the Index S2 isn't really meant for athletes (which is strange, considering it's usually athletes who buy all their other products), and that maybe the algorithm actually works well for people who are very overweight. Maybe that's primarily who's buying this scale? I don't know. But I think the scale should at least be able to pull our fitness data from Connect to be able to distinguish between a morbidly obese person and a relatively fit person.

The scale knows I'm 5'10" and obviously knows I'm ~162 pounds, it should also know that I'm pretty active, so given that information, it should reasonably deduce that probably I'm not carrying 22% body fat.

Again, I'm not expecting medical-grade accuracy, but it would be great if it was at least accurate within maybe +/- 2% Body Fat. An overestimation by 10% Body Fat or more is not just a poor estimation, it's indicative of an algorithm that's just plain wrong, and should be fixable.