But that's not how anyone is doing Keto. It's not that easy to eat the massive amounts of fat this would require, which is why most people on Keto lose weight easily.
Also, why do we need mice to tell us this? Consuming large amounts of calories in general can trigger diabetes, right? In addition, Keto is extremely popular, so why not test on humans?
Hell, even medical students "on average only received 23.9 contact hours of nutrition instruction during medical school" 
Uh...People on keto don't get 99% of calories from fat. There's a significant amount of protein mixed in there.
I don't even know how I would get 99% of my calories from fat without eating a diet of pure butter and animal fat.
The debt would be a buildup of those components.
I do proper ADF or modified ADF when I'm cutting, and good ol' eat-all-the-time when I'm bulking.