I don’t think comparing Geekbench scores directly between completely different processor architectures tells you that much about real world application performa

I don’t think comparing Geekbench scores directly between completely different processor architectures tells you that much about real world application performance. It’s a synthetic benchmark that may be affected disproportionally by factors that aren’t very significant for ‘normal’ use cases, e.g. specific optimizations (compiler, hardware, etc) that hit the happy flow for some part of the benchmark on one architecture, but not the other. But the main difference is thermals and performance under sustained load. So far none of the Appple SoC’s have been shown to handle sustained workloads at heir highest frequency. The CPU in the most recent iPhones is known to throttle under load quite fast, and nobody but Apple knows how it would perform in an active cooling setup.