The post makes an assumption that US labs did not already possess similar optimization. It's also very possible that they did, but are simply not telling anyone in order to maintain obscene margins on cached read, similar to AWS' absurd pricing on bandwidth.