The article speculates that potentially the beneficial effect is actually from magnesium, the fact that there is a differential between tea and coffee and that

The article speculates that potentially the beneficial effect is actually from magnesium, the fact that there is a differential between tea and coffee and that the greatest observed effect is from decaffeinated coffee would seem to indicate that it is actually something present in larger amounts in coffee, lesser amounts in tea, which is not caffeine. Article presents magnesium as a possibility, pointers as to other sources of magnesium, no comment on whether those particular sources have been shown to reduce Type 2 Diabetes risk also but notes that magnesium itself has been shown to reduce the risk. So, in conclusion, it seems very unlikely that the observed effect is merely because an increase in intake of either of the previous drinks will be concurrent with a decrease in intake of other drinks with a high sugar content (Which indeed would be an extremely obvious indicator of how something might assist in cutting the risk of Type 2 diabetes)