Paul Tozour has put up an entertaining video over at the Game/AI blog on the problem of pathfinding - which could probably otherwise be known as 'Wait a minute, that enemy has wings but is getting hung up on a ledge it's flying above. What?'. Along with the video, he's also written a somewhat lengthy treatise on pathfinding in games using waypoints, which he argues are obsolete — and offers some potential solutions. But what about those who say, 'Well, it worked just fine for us in our last game.' Tozour has this to say:

Look at the big picture. Think 10 or 20 years down the road. In that kind of time frame, do you think your games might have lots of different types of AI-controlled characters with different shapes, sizes, and movement capabilities? Will players have AI henchmen that they expect to be just as intelligent as themselves? Will your game worlds be significantly larger, more complex, and more dynamic than they are today? Will you have huge crowds of AI characters — so many that just using simple steering and obstacle avoidance are no longer adequate to make them coordinate with each other effectively? Will your games have realistic physics and huge amounts of physically-simulated objects, and will players be able to use the physics to mess with the AI characters in every way imaginable? Will players be able to change the game world until it's virtually unrecognizable? Will there be AIs in multiplayer that are expected to pass for human players?
If you're interested in nitty-gritty game design elements and AI in games, the article is definitely worth a read through. Even if you're not terribly tech savvy, it's an interesting piece to spend some time with. Fixing Pathfinding Once and For All [Game/AI]