Not to mention that the physiology of the mind isn't very well known. Physicalism of the mind isn't unanimously accepted in philosophy.
If you want a better comparison, you could compare the mind to what happens within the CPU. Other chips would be other organs/subsystems, connected by veins, nerves and various canals (power and data buses). Then you'd have peripherals (limbs) with their firmware (reflexes), and sensors (eyes). All of those communicate together and have side effects on each other.
In the analogy, the body is the whole computer, and the CPU is a part of it, they aren't two separate entities that run in parallel.
If you want to compare hardware <> software, for humans it'd rather be flesh <> electrical/chemical signals.
You'll notice that almost all interesting behavior happens on the right side of the <>. Flesh/hardware are pretty boring, they mostly limit performance and just have to be in good enough condition to function. Slight alterations of those (burnt transistor, internal bleeding) will bring everything down.
The CPU+motherboard+ram+hard disks is closer to the brain. You could argue that power supply and the speakers, microphone, nic, and webcam are like the energy supply and inputs to the brain though.
Sure, and nonphysicalists believe in magic.
One can certainly quibble over the historical accuracy of this, but I think Stepanov describes a worthwhile principle. My sesnse is that a lot of clinical psychology is still trying to abstract over imaginary brains (often by boiling away all the nuance and ambiguity from some ancient philosopher's musings) instead of the set of real brains.
The line may be hard to draw, but we are probably better off trying to draw it so that we don't take too many unnecessary drugs or have invasive surgeries when there are other solutions...