Coincidentally, I just finished Blindsight several days ago so this question is fresh on my mind and I'm excited to discuss it. You've incentivized me to finally sign up for StackExchange after years of lurking. Bravo.
First off, it bears to be said that Blindsight is a work of speculative fiction, meant to explore the idea that self-awareness is an evolutionary disadvantage - it is not meant to be a rigorous argument in that regard. (Watts explains this in his appendices: "Blindsight is a thought experiment, a game of Just suppose and What if. Nothing more.")
A trait can be an evolutionary advantage in one environment while being an indirect disadvantage (merely useless) or a direct disadvantage in another environment. For any given trait X, asking "Is X an evolutionary disadvantage" requires establishing - or making assumptions about - the environment. Breathing oxygen is fantastic in an oxygen-rich environment. It's not so great if you're living in a sulfur vent on the ocean floor. Watts went to great lengths to establish a hypothetical scenario that could result in the evolution of a non-self-aware species like Rorschach, and furthermore to imagine how such a species could be an apex predator in a environment that is extremely hostile to humans (extrasolar orbit around a hyperjovian body with intense magnetic fields).
This is a bit like imagining a chess variant where the house rules are "Rule 1: Any time you move a piece, punch yourself in the face. Rule 2: If you move diagonally, you're exempt from Rule 1" and then narrating a match-up where bishops are dominant.
Some of Watts's counterarguments are a bit weak. (I don't mean this to be a fatal criticism of the novel; again, it's meant to be speculative.) For example, he spends a little bit of time discussing how chimpanzees fail the Gallup Mirror Test more often than orangutans do. He implies that this could be a sign of regressive evolution, a phenomenon where a species loses a trait when it is no longer beneficial. The classic example is when a darkness-dwelling cave animal loses eyesight.
One of the problems with this argument is that the Gallup Mirror Test is not a perfect proxy for self-awareness. It has an unknown, but non-negligible, chance of a false negative: an organism may opt not to explore the mark they see in the mirror because they just don't have motivation to. (For example, children in Kenya are less likely to pass the Gallup Mirror Test. This isn't because Kenyan children are less self-aware; it's because they're culturally conditioned to consider that a mark that appears on the body may have been placed there by an adult for a good reason. Pigeons will fail the Gallup Mirror Test at first, but can be trained to pass it. While it's possible that this shows the pigeon developing self-awareness due to training, that intuitively seems unlikely. Other animals will fail the Mirror Test, but will otherwise display novel behaviors in front of the mirror. Wikipedia has an excellent description of these species-level variations.) Differences in eyesight, visual processing, and motor control are all confounds for the Gallup Mirror Test.
Watts also explores Benjamin Libet's cerebral readiness potential experiments. When Blindsight was published (2006), the most popular interpretation of the Libet free will experiments is that the brain "chooses" to make a decision before the participant is aware that a choice has been made. According to this interpretation, our decisions arise from deep in our unconscious brain. We do not have "free will" and our conscious selves are just spectators to our own behavior. Our feeling of agency is an illusion. However, in 2012, US-French neuroscientist Aaron Schurger ran experiments that support an alternative explanation of Libet's results. I won't go into too much detail, but you can read about the classic interpretation of Libet here and the Schurger interpretation here. (Note: I'm not a believer in free will and I'm not arguing for the concept. But that's a whole 'nother can of worms.)
Benjamin Libet also had an interesting perspective on human volition that speaks directly to one of the exchanges in the novel. Libet hypothesized that our volition serves an inhibitory function. Perhaps our choices truly are made deep in our unconscious brain, however, our conscious minds filter out unnecessary behaviors. This is conveniently consistent with some observations of humans with traumatic brain injury, where damage to the frontal lobes can cause behavioral disinhibition. The glib formulation of this is: maybe we don't have free will. Maybe we have free won't.
How is this interpretation relevant to Blindsight? Late in the novel (spoilers ahead)...
the Gang of Four has a conversation with the second biologist, Robert Cunningham, about how the brain is an extremely expensive organ in biological terms:
Gang of Four: "Brain's a big glucose hog. Everything it does costs through the nose... So sentience has gotta be good for something, then. Because it's expensive, and if it sucks up energy without doing anything useful then evolution's gonna weed it out just like that."
Cunningham: "Maybe it did... Chimpanzees are smarter than Orangutans, did you know that? Higher encephalisation quotient. Yet they can't always recognize themselves in a mirror. Orangs can."
Imagine for the sake of argument that Libet's hypothesis is correct. That complicated expensive brain then prevents us from engaging in behaviors that are also expensive (running away from imaginary predators in the shadows), or high-risk (running away from imaginary predators in the shadows right into the jaws of a real tiger in the bush).
Maybe Libet is wrong. However, despite Watts's familiarity with Libet's experiments, Watts didn't engage with Libet's interpretation of consciousness. It might be interesting to read what Robert Cunningham thinks of free won't.
Finally, the lynchpin is that the evidence for the evolutionary fitness of the human brain is all around us. Our technology, culture, and language has allowed us to become, for better or worse, the dominant species on our planet. (This doesn't mean that the human brain is perfect, or even particularly good at what it does. Just that it's better than the competition so far.) It's hard to imagine how we might have done that without all of our cognitive abilities, including self-awareness. (That's part of what makes Blindsight such a stimulating work of speculative fiction - it skillfully challenges a ubquitous aspect of our lives.) For regressive evolution to occur, there would have to be a dire alteration in our environment.
Once in a while, a science fiction story posits a predator that feeds on, or is hostile to, consciousness. For a couple of video game examples - undoubtedly inspired by Blindsight - I'm thinking of Mass Effect and 2017's Prey, respectively. I'm sure there are other science fiction novels with similar themes. In the face of an overwhelming adversary that attacks consciousness, then perhaps regression to a simpler state of cognition might allow us to carve out a niche and continue to breed.
But at the moment, such monsters, much like Blindsight's Rorschach, remain imaginary.