The problem stated in this particular question can be addressed from software testing perspective, e.g., a hardware can be treated as black-boxed function taking an input and returning an output. In this sense, I can see no difference from, let's say, testing a software deployed on mobile phones. Are we testing then: a hardware (a mobile) or a software? I think integration of both. Going further, providing clear distinction where the testing of software ends and testing of hardware starts can be hard if we start talking about testing circuits with embedded software installed or a firmware in my digital photo camera.
Therefore, I believe each such case should be discussed separately and one that flags it as off-topic should at least give a hint what kind of issues may appear in testing of such a thing that are peculiar to hardware, and thus would not be observable when testing the software in isolation or on a simulated hardware. Explaining such differences would be more informative for the software testing community than ignoring such questions at all.
This, however, would require a larger sample of hardware-related questions. For now, the number of posts tagged with
hardware is very low. Maybe, we should start tagging more questions with this tag, if it seems related? For instance, some errors appearing in the application (like in this question, not tagged with
hardware yet) may be either a software or hardware fault, and it would be valuable for a software tester to understand how to perform root cause analysis in such cases.