This isn't so much about modeling thought processes as it is about illustrating how even in a simplified model one of our debugging approaches fails.
The logic that they're arguing appears to be:
"If we can't even properly reverse engineer an extremely simple deterministic computer chip using fault modeling, it's extremely unlikely that we can infer the mechanisms of an extremely complex non-deterministic processor like the brain."
I do wonder at what level the reverse engineering is done. Also I wonder if their method was pure enough to initially consider the 6502 to be analog rather than digital. That would be a nice trip down the garden path right from the get go.
Now I would say that many fields of study at the higher levels, such as economics, medicine, etc. etc., are incomplete. There's a lot still to be learned. And taking a sidestep of looking at an artificial "brain" from a neuroscience perspective is a good way to navel gaze, fix up your own thought processes. Learning about learning often involves testing your harebrained theories on something not tailored to your experience. Especially neuroscience is the ultimate in learning about learning, if it can be believed that the biological mechanism that does the most learning is made up of neurons networked together, the good ol' neural net.
So to the question of how deep the reverse engineering went. If you want to study the brain, you want to look at low and high levels. The quantum level is the lowest that one probably has to go, and the highest level may involve different groups of minds (psychology of the masses). Computers and brains can be considered at such low and high levels analogously. The first time a neuroscientist looks at an electronic brain, the depth of analysis would probably eschew the quantum level as being too onerous.
I wonder how well neuroscience can explain the functioning of a transistor from a reverse engineering standpoint. The forward engineering of a transistor is to use it for a particular purpose. The neuroscientist would have to calculate what that purpose is. Certainly an engineer would have to be able to calculate whether the transistor actual serves the purpose and does not misbehave. The engineer would design a circuit that is amenable to such a calculation. All anyone neuroscientist or not has to do is to acquire the electrical characteristics of each circuit element (transistors, resistors, capacitors, inductors, power, conductors) and then apply equations that are derived from energy balances and electromagnetism. That should be within the realm of someone who studies brains. Quite possibly brains involve even more complex physical phenomena at the quantum level. Even the complexities of the chemical level outstrip the complexities of electrical circuits, particularly compared to the 6502. All the same, someone who knows physics and who even is given the full knowledge about the physical structure of a 6502 would see that calculating the behavior of this system would be a bit of a job. Chip designers themselves require massive computing to derive that their creations will work exactly.
Paradoxically, this exacting might be the neuroscientists' undoing. Brains aren't exact. Brains fart (malfunction). Perhaps it is enough for a neuroscientist to reverse engineer an adder circuit, a bit shift circuit, a memory writer circuit, etc., and then to determine a handful of microinstructions.
Now I wonder, reflect this analysis back to the brain. I don't know much about this. Are there microinstructions in our heads?