>>While you could use a bathroom scale that pretends to measure body composition, the measurement technique behind it is as unreliable as reading tea leaves.

No argument that DEXA is the 'gold standard, but some 'body composition' scales seem to do a fairly decent job of reporting day-to-day changes in the same individual. For anyone wondering, they use something called 'bioelectrical impedance analysis' (BIA), good summary here: https://pmc.ncbi.nlm.nih.gov/articles/PMC4118362/

I'd consider some of the better scales (multi frequency, multi-segmental, and FDA registered) to be appropriate for tracking day-to-day progress - although BIA becomes progressively less accurate as fat-percentage goes up, so there's that.

I think they'd ought to be in PCP offices as a screening device, not so much for people who are obese (obvious to a PCP) but the ones who look normal but have low muscle mass (AKA "skinny-fat") - if only as a way to persuade them to get a DEXA.

That said, if you consider what T2DM is (high fasting glucose) and the likely culprit (inflamed adipocytes feeding glycerol to the liver (conversion to glucose is substrate controlled), reducing fat mass is a no-brainer, either surgically or now via semaglutide.

And finally, I've seen some research linking long-term microbiome changes to yo-yo dieting, and other research (in mice) linking the microbiome to obesity. I can't imagine that "Frankenstein yo-yo dieting" improves the situation. Bright side: we have a natural experiment happening right now.