I measured a peak bandwidth load of 25 Mbps over Wi-Fi with 32 GB installed and just 5 Mbps with 64 GB.
I’m questioning his expertise if he just accepts that more RAM results in 80% less data downloaded. That doesn’t hold up to any scrutiny whatsoever.
Sounds like things might’ve been cached from a previous test, and if he’s not making any effort to rule that out I wouldn’t trust any of his conclusions.