I might be missing something, but at the moment I am unable to appreciate this article in any way. It creates a ML model that identifies glucose as main factor to diagnose diabetes when high sugar is actually the definition of diabetes.
On the other hand it does not recognize that not all diabetes is the same; pregnancies are probably a factor because of gestational diabetes, the BMI relation is basically the opposite in type 1 diabetics, type 1 starts an earlier age etc. Also I don’t think diabetics will appreciate your cavalier attitude when you say that this “is not the worst thing to have” (long term damage from diabetes can be pretty gruesome), especially type one diabetics for whom self-management and self-control will do nothing; they need insulin.
It would seem that any action based on this model would hurt those who do not suffer from type two diabetes (also I think quotes from wikipedia should be acknowledged), so I would very much like to be convinced that my worries are unfounded.