It’s not as simple as just “faster” suggests (but marketing likes to put things simple).
There exists strong bias on what benchmarks you select. Benchmarks focus on different computing workloads and a processor shown to be good in one is not necessarily good in another. Ideally you sell your product by choosing a benchmark that makes it look good.
Then there is the issue of compatibility. SketchUp 2020 hasn’t been released for M1 and it’s not listed in the system requirements. I would not consider it to be save for presentations or production-ready. Since SketchUp 2020 is running in an emulator and did not receive any touches for M1 from SketchUp developers (it was released before Apple’s dev-kits), the experience is in the sole responsibility of how good Apple’s emulator is.
Then there is the issue of optimization. While the original paradigms used by the Intel architecture have several downsides, it is optimized to the limits of what is possible (complex instruction set, but mitigated internally using clever pipelining and the opposed reduced instruction set). Math libraries and computational frameworks have had decades of time to become optimized for Intel processors. Software using these frameworks can benefit out-of-the box from that speed.
Now Apple’s new architecture may be the better hardware, but the ecosystem can not fully exploit it for years to come. Yes, you hear positive reports for selected applications in the emulator, as well as popular ported applications. But a port makes an application just compatible. If it relies on other libraries and years of optimizations, it might take again years to achieve that level of optimization. As an example, some workloads like in data science actually do not perform well on M1.