I think its It's multivariate conditional. The sweetener doesn't provide direct calorific input, but that encourages more to be drunk and so exacerbates the problem. Weight for weight liquids taken, to energy, to signal received by endocrine system, to outcome on blood sugar levels.
Now add genetic variability to risk/predisposition, exercise levels, bmi, gender, age, co-morbidities..
There is nothing in the article or studies to justify your point about encouragement. The points raised at the start of the thread are still very valid - who participated in the study and whether there were other important factors which were negleted intentionally or not.
Anecdotal point - I'm fit, it doesn't matter if I drink sugar free soda or water. My calorie intake doesn't change because of that at all. It changes due to other factors like lack of sleep or stress levels.
Now add genetic variability to risk/predisposition, exercise levels, bmi, gender, age, co-morbidities..