Being a p-zombie means to do function, without consciousness. 99% of our brain processing is unconscious, so we are p-zombies for almost everything we do.
Chalmers limits his p-zombie argument to all or nothing p-zombies, not "mostly p-zombies". The vast majority of philosphers do reject the plausibility of the actual existence of all-p-zombiehood, and most claim that an all-p-zombie is not possible.
I have read extensively in the philsoophy of mind, but have not yet found any of the rationales that justify "all-p-zombies are impossible". The only justificaiton I have seen is a weak/fallacious one, of ASSUMING an identity theory.
I call this a fallacy, and reference our mostly p-zombiehood to suppport this. As we are mostly p-zombies, the identity of consciousness with either function, or with neural net processing, is pretty clearly falsified. This is a different use of the p-zombie thought concept than Chalmers uses it for, and it leads to different conclusion than Chalmers reaches.
What our partial p-zombiehood reveals is that there is something evolutionarily critical that consciousness does for us. And this is a key fact that any model of mind has to expalin, as a first order observation. IE, functionally, somethings we do go causally through consciousness, and there are aspects of ourselves that prevent these things from ever being done unconsciously. There are two models of mind with predict this -- emergent causally independent dualism, and ensoulment spiritual dualism. No physicalist or purely functionalist model of mind can predict partial p-zombiehood.
Daniel Dennett, Susan Blackmore, and Paul Churchland all reject p-zombies as incoherent, yet their models of mind hold consciousness to be a delusion-- IE we are p-zombies. When the advocates of all-p-zombiehood deny the validity of all-p-zombiehood, this makes me suspect the integrity of their rejection of the concept.