As a homeschooler raised ~20 years ago, the key insight is that outcomes are bimodally distributed based on an overlayed function of parental socioeconomic status and student talent.

The question is which causes which. Does homeschool follow or bolster student talent. Is high parental involvement a meaningless correlation of some other aspect of high SES which actually explains these results, or is it that high SES homes are more involved, and the money itself is the mere correlate?