I guess I have a problem with applying a "one size fits all" policy to device testing. If this thing were going to be implanted in my belly and dispense drugs, I'd want human testing. The bioequivalence testing that MNKD performed more than satisfies most common sense thinking about the device's function, i.e. deliver the same amount of the same drug to the lung. Apply an ounce more of common sense to its use, convenience, disposable status, and you have a very good reason to tell anyone challenging the approach to go jump in a lake, including Schmuckrelly or whatever his name is. The FDA is spineless. They were afraid of getting sued by this guy so they issue a CRL. I would have taken the quoted policy and submitted a revision request and defined a device at least in a couple of different ways to allow for distinction in how they need to be tested. Just saying.