In lieu of not having an AI dragon though, shouldn't the strategic and tactical AI prepare for it? Strategically selling Wizard towers not constructing them, (or anything to do with Dragon cards), I'm assuming AI doesn't research or spend research on Dragon abilities. And its build order in the RTS doesn't seem to include anti-air all that often.
Not sure which direction they want to take the Dragon Commander route in regards to AI its difficult to make those decisions.