I was doing some background research on audio recording techniques and found this discussion of how recorded music evolved over the last twenty years with the adoption of CDs and changes in how music is mastered. The graphic below, clipped from Chicago Mastering Service's page on loudness, caps a series of comparisons between older and newer recordings, of which the recent remastering of Iggy Pop's Search and Destroy is just one (pretty vivid) example:

For point of comparison, CMS also has waveforms for My Bloody Valentine's 1990 Only Shallow (pretty boxed up, but w/ a highest average RMS of -17.3 dBFS and max peak of -4.2 dBFS), Nirvana's 1993 Heart Shaped Box (less boxed up, but louder—highest average RMS -12.7 and max peak of -0.2), and Radiohead's 2001 Dollars and Cents (-6.3 and -0.09). I've been listening to MBV recently, and the graphic explains two things: The overall recording is very loud and boxed up (intentionally so) but it's also very quiet, overall, compared to other recordings. More recent recordings tend to want to fill the sonic space as much as possible.

Obviously, these are only samples from a wide range of recordings, but the tendency is pretty easily observable across popular music. Mastering engineerings (and a lot of musicians) hate this, since the practice erases differences in volume from one part of the song to another. Someone—a lot of someones—seem to like it a lot.