Look at Dark Rock Pro 4 result - it's way too poor considering the size and technology. How did that happen?
Dark Rock Slim, on the other hand, looks very overrated here.
What if there's a trend? What if the test system is creating less heat than it used to? Intel security fixes maybe?
Truth is: neither TPU nor eteknix test coolers properly. You can't just plug it in and run some benchmarks. These results make little sense for a reader. You can only compare the tested coolers and that's based on assumption the system hasn't changed.
You can check silentpcreview to learn how cooler tests should look.
At very least: the base should be fan speed, not CPU load. There's just way too much error in how "general PC" sites do this.