I wonder if there is room for a product, a kind of browser-in-a-website, that would eat those big-ass webpages (server-side) and spit out just the text and (hea

I wonder if there is room for a product, a kind of browser-in-a-website, that would eat those big-ass webpages (server-side) and spit out just the text and (heavily compressed) jpegs. With a little layout to match the original website. Something like how streaming services adaptively subsample data, or like how NX tries to compress the X window protocol. Obviously this would be patchy, but it could be much better than "FAIL".