I agree with that as well. My point was the more limited one that showing "when I poke at the hardware the observed behavior changes" isn't particularly surpris

I agree with that as well. My point was the more limited one that showing "when I poke at the hardware the observed behavior changes" isn't particularly surprising, or on its own evidence of a uniquely "hardware" etiology for the behavior, since any behavior can be modulated by poking at the hardware, if you figure out exactly how to poke. In other words, it does successfully show that the behavior is "hardware-mediated", but then all behavior is hardware-mediated.