In the "good old days" you could at least count on a verse, chorus, different verse, chorus, etc. structure. Then at some point they were doing "chorus, chorus, chorus, bridge, chorus...". Soon we'll be lucky to get more than a repeated sentence fragment.
Case in point: https://www.google.ca/search?q=turn+down+for+what+lyrics
Or this gem: https://www.google.ca/search?q=watch+me+whip+lyrics
A better example of a pop song might be Twist and Shout; not much variety there either.
Latin church hymns, for example, largely use the same corpus of Latin for lyrics, but no one is about to accuse Mozart or Beethoven of a crisis of originality in lyrics.
Case in point: https://www.cpdl.org/wiki/images/3/32/Mozart_427_Great_Mass_...
That's not a good metric.
If I looked for a good metric, I would probably take min(word per meaning and touch), but that would raise a question on what ‘meaning and touch’ is. Nonetheless, good lyrics always seem to be un/under-spoken or sharp cut.
There may be an objective sense in which a song from thousands of years ago has better lyrics than one from today, but nobody can perceive it through the language barrier.
Music that combines real instruments and electronic instruments can produce a much wider variety of sounds than either type alone.
And electronic effects are hardly a new thing - they've been around since the mid 80s.