There might be some interesting observations in here, but this really feels like a grab-bag of complaints, many tangentially-related at best. The author doesn't really do any work to synthesize them to support his central claim.
This is buried toward the end of the article but I think it's a big statement with massive consequences for many practical issues.
Another explanation: those who hyperoptimise for efficiency get what they pay for.