Hi the right er...,
Really interesting question and I'm glad you're finding treatment useful.
A little bit about myself, as I have some knowledge in this area, I've got an Msc in clinical neuroscience and am a medical student; I've also read most of the literature out there on schizophrenia and neuroleptics.
Anyway, first off the dopamine hypothesis is far from proven, and its claims are based upon the so called effectiveness of drug actions, which I'll come to in a moment. You need to be careful when putting a lot of faith in it as a pathogenesis of schizophrenia. There is lots of work with glutamate and serotonin receptors in schizophrenia these days (not to mention all the psychological explanations), and to be frank the dopamine hypothesis is old news. It is a theory that largely became accepted belief without ever really getting strong evidence behind it.
On to the drugs themselves. Typical first generation antipsychotics e.g. chlorpromazine were discovered originally as anaesthetic drugs and were first labelled as major sedative hypnotics i.e. tranquilisers. If you give anyone a D2 receptor antagonist they will calm down; this effect is obviously just more pronounced in schizophrenic patients. They are general sedatives and are far from a targeted cure for schizophrenia that they are often portrayed as. D2 receptors are in large quantities in the basal ganglia (a group of nuclei involved in movement) and are involved in movement pathways, which is why their modulation often leads to movement disorders seen in antipsychotic treatment (tardive dyskinesia) and why their loss in Parkinson's disease has such profound effects on movement.
On the receptor issue you are very right that receptor density changes in response to transmitter supply, hence why they increase when you deprive the brain of dopamine.For this one you need to think about genetics though. Receptors are just proteins implanted in a cell membrane and encoded for by a gene. If the fault does lie within the receptor it is because you have a genetic abnormality, so all subsequent receptors of that particular type would be translated as defective.
The only way you would suddenly get healthy receptors being produced is if there was an activation/inactivation of genes through some means e.g. epigenetic mechanisms (an explanation best left for someone who knows far more about genetics than I!). So it is unlikely that you would get widespread mixing of healthy and defective receptors. Effectively the receptors would either all be abnormal or all be normal, unless something altered the gene e.g. radiation or epigenetic effects etc etc.
Interesting question!