Why isn’t quadtree compression widely used for lossy image formats?

One of the key issues with quadtree compression is that it's computationally intensive. You're increasing the number of processes for loading images, which can slow things down considerably. Plus, the potential blockiness can detract from the image quality, which is why many might choose lossy formats that offer better visual fidelity for the trade-off on file size.

Quadtree compression isn't as easily parallelizable as other methods, which can limit its efficiency in processing. This can be a big factor in why major image formats haven't adopted it widely. The ability to streamline the compression process is crucial, especially when dealing with high-resolution images.

Quadtree compression has been around since at least 1992 and there are ongoing discussions in the field. While it can perform admirably in certain situations, algorithms based on DCT or wavelet techniques tend to yield better peak signal-to-noise ratio (PSNR) for a variety of image types. Typically, these methods are also efficient for both compression and decompression, plus they easily integrate with hardware for speed improvements. This makes them more favorable for general use.

Your idea seems perfect for images with large blocks of solid colors, like logos or vector graphics, but it might struggle with intricate photos that have a lot of variation. While it could work well in niche areas, it faces stiff competition from existing solutions like vector images, which already handle those types of graphics quickly and efficiently. In a way, there are already better solutions out there for what's essentially a similar need.

The problem with relying on quadtree compression for photos is that these types of images rarely have large, uniform color areas. Even images that appear solid often have gradients and textures that aren't captured well by simply averaging a block of pixels. Lossy compression tends to skew the nuances of these images, where lossless methods might actually do a better job for such visuals.