Here's a better source (PDF), link-chained from yours.

On brain changes due to drug use:

Medication-Matched Subjects. To address the possibility that neuroleptic exposure and/or lower IQ could have determined differential gray matter loss in the schizophrenics, we mapped 10 serially imaged subjects referred to the childhood schizophrenia study who did not meet diagnostic criteria for schizophrenia (labeled Psychosis NOS - Not Otherwise Specified - in DSM terms; (24)). These subjects received identical medication to the patients in this study through adolescence, primarily for control of aggressive outbursts, and at follow-up, none had progressed to schizophrenia (35) but all continued to exhibit chronic mood and behavior disturbance. While medication is unlikely to be responsible for a loss profile that moves across the brain, clozapine, for example, may increase Fos-immunoreactivity in the thalamus (36), and might, logically, modulate rates of cortical change. (In addition, brain regions important for motor function, including the basal ganglia, show increased volumes in response to some older, conventional neuroleptics, although these effects are renormalized after treatment with the atypical antipsychotics used in this study). As seen in Figure 6, while the non-schizophrenic group did show some subtle but significant tissue loss, this was much less marked than for the schizophrenics. Moreover, no temporal lobe deficits were observed in the PNOS group (Fig. 6), suggesting that the wave of disease progression into temporal cortices may be specific to schizophrenia, regardless of medication, and also regardless of gender or IQ. Intriguingly, the psychosis NOS subjects, who share some of the deficit symptoms but do not satisfy criteria for schizophrenia, exhibited significantly accelerated gray matter loss in frontal cortices relative to healthy controls, in approximately the same, but a less pervasive, region than schizophrenics (a significant loss of 1.9%±0.7%/yr. was detected in both left and right superior frontal gyri; p<0.03).

So the answer to my question appears to be that drugs may or may not be doing some brain damage, but not nearly as much as the whole change seen in schizophrenia.