It does not counter his argument, but this doesn't mean he's right, either. His argument states that given a platform function, developers have historically us

It does not counter his argument, but this doesn't mean he's right, either. His argument states that given a platform function, developers have historically used/exploited said function in ways that its creator had not intended. In terms of device functionality, this comes in terms of new applications for and of that device. The OP surmises that increasing the number of "opened up" functions via providing an API for developers will increase the usefulness of the device via the previous statement. Following his reasoning, the ultimate functional and useful device is "open" in the sense of FOSS because developers have the ability to exploit all functions of the device. My question to that conclusion would be, what if hackers made changes to the device that ran counter to the initial design philosophy of the device...which is what the users initially purchased the device? Is this a good thing always? Is this a bad thing always? The answer is neither, and this is why the implied (as interpreted by myself) conclusion is false. The OP's argument doesn't necessarily follow because usefulness does not equate to functionality and hack-ability. Good design is not something you can quantify by "featurefulness". If it were, users would have stopped buying new copies of Microsoft Office years ago. Following from the end line of Paul Graham's post, if you give hackers an inch, they will take you a mile. PG doesn't state whether or not that is a mile in the direction your preferred direction of travel.