"you get over 20 pages of large thumbnails all on the same huge page....And because thousands of useless large thumbnails are loaded, a huge waste of user bandwidth takes place each time."
Beware commenting on things you don't understand, because often it just makes you look like a fool.
In this case, Dvorak apparently didn't notice that while all of the images "were on" the same page, they were simply placeholders until you scroll.
If an image exists outside of the viewport, does it exist? Not necessarily.
As to them being large images....welcome to the high speed world, John. We've graduated beyond dial-up.
"So instead of being able to scan the entire page with a simple glance to find an image from say, usgs.gov, you have to put the cursor on each and every image and wait for a pop out with the information."
Who searches for images by domain like that? Few users of Google image search, I suspect, with the average user casually looking for some images to "borrow" for their book report or blog. The source domain just doesn't really matter to most users, again speculating (though presumably Google has actually verified this).
But say you really want to see images from usgs.gov? site:usgs.gov, e.g. http://images.google.com/images?hl=en&q=site%3Ausgs.gov+...;
And of course you can search by size (larger than, or a specific absolute size), and so on.
There will always (always) be people who complain about change. John, in this case, is just plying his trollish trade, however.