3d "steps" refers to whether you can see individual levels.
-thinks- I don't have an image so I'll describe it like this. Pretend there is a room with 10 objects in it... This room is being rendered on display in a "stepped" situation with 3 levels. 1 foreground, one background, and one in between. Those 10 items would thus have to appear very close (foreground), very far (background), or directly in-between the two.
This wouldn't be so bad if all the items were either right on the camera, all the way at the end, and directly in the middle... but most of the time, items thrown about the room won't be so orderly and fit on "levels". You'll have some items that are much closer to the front than middle and vise versa... In a stepped system though it doesn't matter where they are in the middle of the room, they'd all look like they were exactly in the middle do to the limited number of "steps"
Another (simpler) way to describe it... A step-less system is one that doesn't have "layers" of 3d, it appears natural and objects have true depth and roundness... Stepped or layered 3d is more a collection of flat 2d surfaces placed in rows. Think like the original Doom game. There were no 3d models, but it gave off a sense of depth because the 2d images moved in a 3d plane.