According to the above 1977 study, acetazolamide therapy in Hypokalemic Periodic Paralysis patients resulted in less of a drop in serum potassium during an attack due to reduced glucose and insulin levels than those who did not take acetazolamide. It still drops when an attack is induced, but not as severely. The higher the glucose, the higher the insulin response and the harder potassium shifts as a result. In a nutshell, less insulin response = less potassium shifting. That would explain why there is less paralysis.
It makes sense now. Brilliant.