Well, the idea is fairly obvious - if you read the Wikipedia page on diabetes, it says that glucose in urine is caused by the reabsorption mechanism being overwhelmed; lowering the threshold and excreting excess glucose seems worth considering if the alternative is storing it all as fat.
Of course, that's just a theory and practical effectiveness and safety are quite a different matter, which drives up the cost. The reason patent drugs are so expensive (compared to generics once the patent expires) is because drug development is expensive, and pharma companies need to recoup the cost of development.
[In theory, anyway. In practise, they are just as likely to make minor alterations to drugs to extend the patent and get rich]