The encoder currently isn't using a lot of format features like curves or layers (which would partially also require deeper integration in for example Photoshop

The encoder currently isn't using a lot of format features like curves or layers (which would partially also require deeper integration in for example Photoshop to pass such information to the encoder).