In schizophrenia, this process goes overboard and the brain compensates by treating random noise as signals, which leads to hallucinations. People with autism apparently don't trim enough which impairs learning. Thus, it could be that both conditions involve defects in the same process in opposite directions.
These findings may suggest new treatments targeting GABA receptors for "normalizing" synaptic pruning in diseases such as autism and schizophrenia, where synaptic pruning is abnormal. Research has suggested that children with autism may have an over-abundance of synapses in some parts of the brain. Other research suggests that prefrontal brain areas in persons with schizophrenia have fewer neural connections than the brains of those who do not have the condition.From here.
On the other hand, it seems odd that autism would have a very early childhood onset if it is really related to synaptic pruning. But, there are also other developmental periods besides late puberty when synaptic pruning occurs, and perhaps an earlier time period is implicated in autism.