As everyone else has said, I would prefer that (if I were browsing with no filtering).
Here's the trivial client blocking method: OCR and break the result into different groups by page location, font size and color; the article will be the largest groups of text with the same font. Large images and charts that were part of the original article would be cropped out and re-inserted into the final document. Within a few days you would see a browser extension made with adaptable filtering for most sites that used the monolithic PNG.
Here's the trivial client blocking method: OCR and break the result into different groups by page location, font size and color; the article will be the largest groups of text with the same font. Large images and charts that were part of the original article would be cropped out and re-inserted into the final document. Within a few days you would see a browser extension made with adaptable filtering for most sites that used the monolithic PNG.