Given that the chip in macbooks will be a lot faster than the A12Z, it could very well offer twice the multicore performance.
Now try some math workload on A12Z, i.e. real-world stuff and weep. Most of the die space on x64 CPUs is spent on caches and SSE, so if you cut them out to be more power-friendly, you lose performance in high-end applications.
It's not like anything faster than Pentium J is needed for Office, browsing or even front-end development, so in that segment their first gen should be good enough and maybe even better than i3/i5.
However, it will be interesting to see if that A14 big core implements the ARMv9 instruction set, which is said to be imminent. Apple did implement the ARMv8 instruction set much earlier than it's competitors, so there is a bit of a history there.
ARM teased new features like an up to 2048 bit SIMD implementation with SVE2 and Transactional Memory for the ARMv9 instruction set back in early 2019, but has mostly been quiet.
https://community.arm.com/developer/ip-products/processors/b...