I’m a huge fan of Counter Apologist; check out his blog if you haven’t already.
He recently had a back and forth with Christian apologist Randal Rauser over the knowledge argument, which is a much discussed argument in the philosophy of mind. Here I will offer possible ways a physicalist can respond to the argument, as well as offer a few thoughts on the discussion between CA and Randal.
CA seems more interested in defending a reductive physicalist view; of course there are non-reductive physicalist views, such as token-physicalism and property dualism, which I will talk about here, that are able to get around the knowledge argument.
Possible Responses to the Knowledge Argument
1) Mary didn’t gain propositional knowledge. That is, if we distinguish between “know-that” and “know-how,” it could be that Mary only gained a familiarity with an experience, rather than a propositional fact about the world. In other words, the mere experience of tasting cinnamon or seeing red is not itself a fact over and above the physical facts; learning about it through an experience is only to become acquainted in a new way with a set of facts that one already knew about.
2) The physicalist can defend a posteriori physicalism, rather than a priroi physicalism. That is, the physicalist doctrine can still be true–that the entire physical nature of the world entails the entire nature of the world–but knowledge of this truth is only gained through experience, not a priori. So the physicalist can agree that Mary learned a new fact upon tasting cinnamon or seeing red, but only because she was blocked from “knowing all the facts” while being trapped in the room.
3) The physicalist can distinguish between a “theory-conception” of the physical and an “object-conception” of the physical. (Call these “t-physicalism” and “o-physicalism.”)
T-physicalism defines “physical” as: a property is physical iff it either is the sort of property that physical theory tells us about or else is a property which metaphysically (or logically) supervenes on the sort of property that physical theory tells us about.
O-physicalism defines “physical” as: A property is physical iff: it either is the sort of property required by a complete account of the intrinsic nature of paradigmatic physical objects and their constituents or else is a property which metaphysically (or logically) supervenes on the sort of property required by a complete account of the intrinsic nature of paradigmatic physical objects and their constituents. 
So, if Mary knew everything physical that there was to know about people, on the t-physicalist conception of “physical facts,” it’s plausible to think that Mary did learn something significantly new about the world upon tasting cinnamon or seeing red. However, on the o-physicalist conception of “physical facts,” Mary knew everything about how the universe is structured, all the way down to the most minute details, as well as what sorts of facts/properties supervene on them. And so, she would have know when other people are tasting cinnamon or seeing red without having to ask them, just by looking at their particles. The important difference with o-physicalism is this: the experience of “tasting” or “seeing” is not intrinsic to cinnamon or roses, and is not intrinsic to “cinnamonness” or “redness” either. These are extrinsic, or relational properties/facts, which o-physicalism doesn’t say anything about.
So since o-physicalism claims that physical properties are those that are, or supervene on, properties needed to give a complete account of a thing’s intrinsic nature, and since “what it’s like” properties are extrinsic to red things and redness itself (since they depend on a relationship these things bear to an agent), we therefore don’t need to appeal to “what it’s like” properties to give a complete account of the intrinsic nature of physical objects or properties.
That is to say, “qualia” facts are extrinsic to the nature of the things that we talk about, and so Mary saying “So THAT’S what it’s like to taste cinnamon (or see red)” is no slight against o-physicalism since those facts are not needed for a complete account of the intrinsic nature of anything (except for qualia, but the existence of qualia is exactly what’s being disputed so we can’t assume that they should be given an account of their alleged intrinsic nature). 
I agree with The Thinker’s point against Randal’s view of substance dualism and agent causation:
…given everything we know about physics – particularly about the atoms that make up our very material brains it effectively rules out the idea that there’s something else out there that can affect our brains without violating the known laws of physics. Given that said laws are extremely well established, we should be very skeptical of things which claim to violate them.
Indeed, the success of science on the assumption of the causal closure of the physical is a powerful reason to reject Randal’s view. This is, however, more closely related to mental causation than it is to the existence of qualia and the question of non-physical facts.
Second, CA quotes Randal as saying:
If a thought/sensation just is a pattern of neurons firing then every property of the thought/sensation is a property of the pattern of neurons firing.
The view that Randal is referring to here is “type-identity” theory, where every mental property is identical with a physical property. This entails supervenience physicalism; however it is important to point out that this is not something all physicalist’s must be committed to, and many are not. In other words, don’t let the anti-physicalist frame the knowledge argument such that the only possible response must be from a type-identity theorist.
It’s also worth noting how token-physicalism differs from type-physicalism. Token-physicalism says that for every particular object, event, or process x, there is some particular physical y, such that x=y. Notice, however, this view leaves open the possibility that some objects, events, or processes also have contingent mental properties (in addition to their physical properties). So in this sense, token-physicalism does not logically entail supervenience physicalism or type-physicalism. But it can still be considered an “identity” theory of physicalism.
There are also other monistic views, such as Russellian monism, that can answer the knowledge argument. These, however, require a re-definition of “physicalism” to either include mental or proto-mental qualities intrinsic to physical substance, or have the physical and the mental be two ways in which a third “neutral” substance manifests itself in the world.
The physicalist has many ways to respond to the knowledge argument, and “qualia” is not some knock down anti-physicalist fact of the world.
1. See here for more: https://plato.stanford.edu/entries/physicalism/#TokTypPhy
2. Paraphrased from a correspondence with Dr. McElhoes at Arizona State University.
SEP on qualia: https://plato.stanford.edu/entries/qualia-knowledge/