At first I wasn’t very impressed with this magnifying glass—I mean, who wants to view a pixelated portion of an image? Then I viewed the large, higher-density collage further down the page. When linking in and sizing down a larger picture, the magnified portion doesn’t become pixelated until you get to much higher magnifications.
I can imagine this being useful for maps and diagrams where both the overall layout and details are relevant. I’ve been thinking about large-scale software maps for years now—ever since I connected Isaac Asimov’s “Psychohistory” with flow charts and data flow diagrams.
This technique isn’t really the answer, since an image large enough to be truly useful would be slow to load and difficult to create and update. Instead, I’d like to see a similar technique applied to SVG.
In the mean time, it might be interesting to combine a Google Maps (tiled) approach with the magnifying glass technique, although the current code likely doesn’t support magnifying pieces of adjacent images.