Charle's argument seems to be that peer review somehow ensures the quality of your idea or solution. The problem is, there is significant evidence (ironically often in the form of peer reviewed papers) that this is not the case:
http://www.lutz-bornmann.de/icons/BornmannARIST.pdf
https://www.scientificamerican.com/article/an-epidemic-of-false-claims/
so they... shouldn't try so hard? And just release rushed buggy hackable code that may have far reaching consequences not easily foreseen. Good point.