Larry Levy has an article on learning curves:
Because there is so much less time devoted to the newer titles, it seems that many gamers are unable or unwilling to look too deeply into games to find their nuances. More and more, it seems that the worst thing a game can have is a learning curve.
He mentions Euphrat and Tigris, but not more recent games like Puerto Rico or Caylus. Let's look at the rankings, shall we? I don't see anything dismissed out of hand.
I think Larry's railing against snap judgements. I happen to have made snap judgements regarding Cleopatra, Augsburg (two games he mentions), Euphrat, Puerto Rico, and Caylus. The latter three were "Good. Interesting. Buy it." My snap judgement for Cleopatra is "Not bad, but nobody would be playing if it were cardboard." I didn't strongly care for Ausburg, but it turns out I had a rule wrong.
Now, it may be that some games are more vulnerable to snap judgements. Those games with hidden glory (and obvious stinkers). I guess they're subtle, which means that they have learning curves by definition. But many games with learning curves are not subtle. I've mentioned before that it took me several games before I could last through 1830 without bankrupting ... and when I taught players (several years later) I won something like 10 striaght games. It has lots of tactics, and a learning curve, but you know exactly what you get on the first game, when the hollow husk of a corporation gets deposited on your doorstep.
It may be that Cleopatra (and Ausburg) are subtle games. It would be tough to tell after one play, by definition. And subtle games don't garner much praise. "Obvious" games with a learning curve seem to be doing fine.
Well, snap judgements on strategies (as compared to a games qualities) are notoriously finicky. I suppose that can impact (if a game has an obvious strategy). My feeling is that most of the people who find these things will eventually realize the error of their ways. But my world is populated by wonderful people.
Like Larry.