Optimizing established functions/code, like QS. There's still a point?

Greetings. Apologies for the title, but it's length is rather limited.

I'm asking, because I've seen the article about google's simd based quicksort, and because I've actually thought about this recently.

I've neglected to pursue this problem, because I was thinking that there's no point to it anymore. I was thinking that everything solved to best efficiency already anyway, so why bother?

Am I wrong? Is there still room for improving efficiency/performance of at least some functions/algorithms out there?

If so ... what should I tackle? To make this easier, just assume I can improve literally anything, as long as there's still room for it, and throw the challenges/problems at me! :D

Thank you!