Vector graphics are superior, but raster graphics will be needed

Vector graphics provide a clear edge over raster ones when it comes to clarity as zoom in and zoom out operations take place; you don't worry anymore about the resolution. However, it's too early to dump raster graphics. 

Usually, raster graphics are associated with resolution; they don't mean much unless you know how many dots per inch (dpi) are there in each pixel and how many pixels are there along both dimensions (in 3D graphics, you are still dealing with two dimensions as long as you watch it on a screen... it's not 3D for real unless, let say, you can see the back of the object by looking from the back and the front by looking from the front, like a doll in your hand, which is impossible with screens). All satellite imagery nowadays are resolution based, QuickBird, LandSat, Ikonos...etc, and all digital cameras give you pictures in raster format. The problem with this format is the size of the files and the clarity. The moment you increase the resolution, the size increases exponentially, not linearly, because it's two dimensions affecting each other. This affect everything relevant: the bandwidth, the storage, the speed...etc. Also, once you zoom in too much, you don't see clearly. It's similar to the optic zoom versus digital zoom in digital cameras and camcorders; digital zoom doesn't mean much at all, while optic zoom is what you need to be high.

When it comes to vector graphics, the concept of resolution vanishes because it's not pixel-based. This allows you to zoom in as much as you wish, and you still see 100% clarity. Whether you see a detail or not depends on the existence of that detail, not the zoom level, while in raster you might very well miss a detail because of the resolution while it's there. Also, in geographic information systems (GIS) the vector format is always giving smaller size when compared to raster format.

The issue of converting raster format to vector format requires a process called digitization, used to be done by digitizers and not mere mouses are used to carry it out. It is, however, unpleasant process that requires time, effort and skill.

Having vector graphics on the web will allow the graphics to be searchable, while raster graphics (while certain technologies allow to search them) are conventionally regarded anti-searchable; I mean the content itself, not the attributes of the graphics.

So, if Google manages to transform web graphics into vector, it will be a breakthrough in the way we deal with the web. As a matter of fact, the appreciation of the 3rd dimension, while it will remain false as it's screen-based, will be intensified with vector graphics, which might take us into the real 3D generation.

This author consults with leading institutions through GLG