Your first point probably isn't actually true. The research that first connected the two showed that across a population sugar consumption up -> diabetes rates

Your first point probably isn't actually true. The research that first connected the two showed that across a population sugar consumption up -> diabetes rates up. But it turns out that when you control for weight the effect of sugar specifically is much much smaller. To the point where if you're a healthy weight your sweet tooth isn't moving the needle.