AT: "a fruitful theoretical path would be to accept (at least initially) the existence of consciousness as an unexplained fundamental concept" 

In other words, accept that we do feel, and that we cannot explain how or why. I agree. It's true, so we might as well accept it.

AT: "much of the content of consciousness/feeling can be distinguished, described, compared, publically represented, and analyzed"

What we feel can be described, and its brain correlates (which are almost certainly also its causes) can be found and analyzed. Reverse-engineering those will explain, functionally and unproblematically, everything we do, and are able to do. But it will not explain how or why any of that functing underlying our behavioral capacities is felt. And although we cannot do anything about that, it is definitely a (profound) explanatory gap.

AT: "The key question [is] "How does the brain create the gloriously varied content of consciousness?"  

That question will not be answered either. We will find out how the brain generates adaptive behavioral capacity, and, given that generating that capacity also happens to feel like something, we will find out the correlates (and probable causes) of those feelings. I don't think we'll have a substantive explanation of how the brain generates feeling, but I think that there will be little doubt that it does; but not being able to explain how the brain generates feeling is the lesser problem: the fact that we cannot explain why (functionally speaking, i.e., causally speaking) the brain generates feeling is the greater problem: all those gloriously varied feelings, when all that was needed for adaptive purposes -- and all there is causal room for -- is the underlying functing. The fact that (some of) those underlying functions happen (for mysterious, unexplained reasons) to be felt just stays the dangler it is.

AT: "specifying putative neuronal mechanisms that can be demonstrated to generate activities in the brain that are analogous [to] feelings"

That is unfortunately just correlates again.

AT: "unlike the smell of a rose, the elementary properties and detailed spatial relationships in our feeling of a triangle can be displayed in an external expression which others can observe and examine" 

I'm afraid I can't agree: The geometric properties of detecting and manipulating triangles are functing, and unproblematic. What it feels like to see or imagine or manipulate a triangle, in contrast, is every bit as problematic as what it feels like to see red. (Lockean primary and secondary properties don't help here.)

-- SH