It might be that there is some well-defined physical connection between the computer and the brain. To return to the crystal analogy - NaCl has structural similarities and behaviour similar to other crystals, which are actually dissimilar to molecular chlorine. But this commonality is well-defined physically.
The problem is that just because
some commonalities are well-defined in terms of physical properties does not mean that they all are. Commonalities can be functional as well as physical. A simple example of that would be a bridge, which can be broadly defined as providing the capacity to cross a river or similar barrier. But there are lots of different types of bridges, and you won't be able to find any
physical commonality among them. What you'll find is a functional commonality.
And, in fact, the capacity to cross a river can also be provided by a ferry, which has even less in common with a suspension bridge than other types of bridges do. If all you're interested is (abstractly) moving goods from point A to point B, you don't even care whether they will be transported by bridge or by ferry. The lack of a physical commonality is irrelevant to the question you are interested in, which is the abstract functional property of transport capacity.
Indeed, there are a number of companies out there that specialize in exactly this. The whole point of a "multimodal" transport company is that it will get your goods from point A to point B without bothering you about the technical details of whether it's shipped by road, rail, ferry, container ship, or air. From the point of view of the customer, what is important is the abstract functional capacity (my goods will get there) that comprises a
commonality, not the detailed physical capacities that are NOT held in common.
Computers, of course, are the same way -- we have at this point probably a hundred different definitions of "computer" that are provably TM-equivalent, even though their physical commonality is entirely different. What's the physical commonality, for example, between a Turing Machine and a Game-of-Life? What aspect of Game-of-Life corresponds to the third input cell on the tape? What aspect of a Turing machine corresponds to the notion of a "blinker"?
Computers are defined functionally, not physically.
And since we have no clear-cut definition of "consciousness," you cannot assert that it must be defined
either functionally or physically. The Strong AI hypothesis is simply that "consciousness" (or "intelligence," if you prefer) is defined functionally instead of physically. Penrose claims that consciousness must be defined physically -- but since he gets the basic computing wrong, the basic neuroscience wrong, and the basic physics wrong, there is no reason to take his claims at all seriously.
