The problem with Chalmers' perspective is that he doesn't have any argument except "nuh-uh".

Being a half-adder isn't a fundamental property of matter in any important sense. There's no hard problem of half-additivity. You assemble matter one way, you get a half-adder. Another way, and you get Archimedes' Screw, which doesn't do half-adding but does lift water smoothly via rotational motion. Another way and you get a person, which doesn't lift water smoothly via rotational motion, but does do consciousness.

We live in a panrotationalwaterliftism universe!

Or maybe we live in a universe with emergent properties, one of which is consciousness. (IIT isn't necessarily a panpsychist perspective, incidentally; you can also view it as speculation on the nature of the mechanism.)

I admire Koch as a neuroscientist, but I don't think he's likely to be on the right track with solving the problem of consciousness simply because we don't have a mechanistic understanding of a lot of high-level features of behavior. What's the mechanism of fear of spiders? What's the mechanism of laugh tracks promoting a mirthful reaction? What's the mechanism of logical inference? What's the mechanism of recognizing your grandmother?

We don't know.

We have all the pieces in place to conclude that it's very likely that it could be known, and in some case have neural correlates to parts of the process, but that's different.

Now, this isn't to say that Chalmers-style dualism must surely be wrong. He just doesn't have a very good argument that it's right. It's effectively an argument of the gaps, saying that a gap cannot be bridged, in an area absolutely chock-full of gaps. I do expect the gaps to be filled in, bit by bit, but I wouldn't put any timeline on finding the physical basis of consciousness in any particular amount of time. We're extremely far from "wow, we know everything and still we can't explain it; it's like magic"!