I don't think the CPU performance is down to optimization. The 4 custom Taishan performance cores having hyper threading is probably why. The Kirin 9000S has 8 cores / 12 threads, which is why the multi-core score is so high.

The GPU drivers definitely aren't ready yet; it can't render Genshin Impact correctly. It's not an ARM Mali reference design; it's Huawei own Maleoon 910. The only things from ARM are the Cortex-A510 efficiency cores, and the instruction set.