I am pretty sure, that the dysfunctional glucose sensing and inappropriate liver glucose release are consequences and complications of diabetes, not the primary causes. Diabetes (Type 2) is primarily caused by insulin resistance combined with progressive beta cell dysfunction.

Therefore, treating the liver to treat diabetes seems .... weird?

How would that work? Do we even have a reliable way of detecting localized places of internal inflammation? The article mentions ultrasound imaging, but this is beginning to sound a lot like a "if all you've got is a hammer" type thing.

What about other localized inflammatory conditions, such as asthma?

It was recently discovered that when listening to music your brain waves sync up to the same frequencies, which I had predicted for a long time. This means the "computational model" (i.e. synapses and neural nets) theory of consciousness is mostly wrong.