What the author here has discovered is a correlation: that heavily compressed/limited music does not sell as well, in the long run, as more dynamic music. He is concluding that limiting/compression leads to poor sales. That is a flawed conclusion that fails to account for all of the confounding factors that may relate to album sales.
Here's another example of a correlation: Old people consume more prune juice than younger people. Older people also die at a much higher rate compared to young people. Therefore, consumption of prune juice leads to death. Make sense? Of course not. The two factors are completely unrelated.
While I don't disagree with the author's preference for more dynamic music, there are far too many confounding factors. Namely:
1. Time bias -- Heavily compressed music is a relatively new phenomenon. The well-selling examples he uses of dynamic music (Rumors, Stairway, White Album) were recorded decades ago and have been available for a long time. The albums that have not sold as many copies were all (**shocker**) recorded in the past decade. Of course they haven't sold as many copies. Only time will tell if today's music has the staying power of these older hits.
2. Quality -- Much of today's popular music is throwaway garbage. There was probably a ton of garbage coming out in the 60s, too. However, most of it is long forgotten. The few great albums of that era (a coming of age era for both rock music and recording techniques) are what we are still talking about today. It is simply unfair to compare today's pop music with a few top-selling albums of yesteryear by great bands, with great songwriting and playing, made in great studios by great producers. What will be the gems of our generation? Again, it is too soon to tell.
3. Technology -- The record industry has been lamenting for years the decline in record sales and blaming it on mp3s and such. Also, recording technology is available to many more people now than it was 40 years ago, meaning lots of people are recording music (at home). The well produced studio album is becoming less common. People also have at their disposal (for better or worse) tools that they may not have the experience to use properly. Whatever the cause, the album format may be on the way out.
There are plenty more confounding factors. The point is, even if you find a relationship between two events, there is not necessarily a causality implied. You have to account for as many variables as possible to adequately test a hypothesis.
Oh, and avoid prune juice. That shit (pun intended) will kill ya!