It's an abstraction over an indirect measurement technique which is subject to many if-but-maybe issues. I'm not saying don't trust it but .. I am saying don't trust it without some alternate cuff/sphygmomanometer to calibrate and be aware even amongst cuffs, arm may be rated "better" than wrist.

I still don't see how a non invasive glucose level works tbh.

(Not a doctor or a medical scientist or an FDA)