Although most designs out there follow a "Top Down / Large to Small" convention, I have this debate with my team often. My design professor put it best: a design only needs 1 (or 2 at most) differentiation to serve its purpose.
In other words, a title is ALREADY sitting on the top row. Therefore, that should be good enough to indicate it's the title. How many changes in the design do you need to make? A title doesn't have to be larger AND bold AND underline AND color AND blinking text AND shadow AND (etc etc). You have to admit, it gets tacky after 1 (or 2) design steps.
Devil's Advocate: people say, it looks balanced. That is true that it "looks balanced" but is that the purpose: to look balanced OR to get the content to the reader without distraction?
IMO, your screen capture may look "flat" because of the same font size. But honestly, the middle of the design (the form fields) get all the attention and that is the purpose of the modal, right?