I had a similar feeling a decade ago, but I now think otherwise. We may have reached the good-enough compression ratio but not the good-enough performance, and

I had a similar feeling a decade ago, but I now think otherwise. We may have reached the good-enough compression ratio but not the good-enough performance, and modern algorithms now rentlessly improve on this. On the other hand lossy compression always tends to exploit the currently available computing power; the ability to continue doing so for decades implies that improvements on lossy compression are more "incremental" than those on lossless compression.