Wrong, a 2D being doesn't see in 2D. It sees in 1D and infers the 2D by parallax, lighting and shade, prior experience, etc..
Though, of course, if you ask one, it would claim that it could visualize 2D quite well, since it would have honed its perception so much that the leap from 1D to 2D would pretty much be instinctual and unconscious.
Much like how we 3D beings imagine that we see 3D, but actually we don't. What we see is just 2D images from which our brains, having internalized the process after much exposure to such images from the eyes, unconsciously and instinctively construct 3D models based on things like parallax, light and shade, prior experience, etc.. (Which may or may not represent reality -- hence the existence of illusions like the Necker cube AKA the tumbling cube, the hollow mask illusion (that appears to be a face turning the other way), etc.).
Consider for a moment, how much our so-called 3D perception is really just manipulating 2D surfaces in disguise, so much so that we have come to identify many objects by their surface, regardless of their actual, internal, 3D contents. Even when we visualize containers, which arguably are more than just their outer 2D surface, we still think of them in terms of 2D -- as two sets of 2D surfaces: an outside set, and an inside set. We never think of, for instance, the wood grain structure of a wooden box, for example. It doesn't exist in our minds, because we are basically thinking in 2D, just with some clever imagination that pulls it just a tad closer to the 3D reality.
Yet, from the vantage point of hypothetical beings who can actually see 3D in its full glory, such as 4D beings with 3D retinas, the inside and outside surfaces of the box are laughably irrelevant -- they are no more than the outlines of the box, barely within notice -- while the 3D wood grain structure would be the primary feature of the box in their sight.
Nonetheless, for what it's worth, our pseudo-3D perception is pretty useful when it comes to visualizing higher dimensional space. It at least gives us a crutch upon which we can, just barely, peer into the world of 4D objects and their amazing geometries, given sufficient training along the right principles. It takes effort, certainly, but it is not insurmountable.