That's a weird graph. It's not saying omnivores are 100% at risk of diabetes, it's just saying that vegans are half as likely to get diabetes than omnivores. I get that this is a pro-vegan post, and I'm not doubting it's benefits in this area, but there are less manipulative ways to display information. A better way would have been to peg omnivores at 7.6% according to the tonstad 2009 study they're citing and make it clear the graph does go higher by making the highest amount 10%. It would be readable but less obviously biased and give more information about actual risk with each diet.

"Compared to 100% of an omnivore's risk" is an interesting way to display this data

I thought this was gonna be comparing diabetes between herbivores, omnivores and carnivores