You didn’t mention what precise CPU you have in your Windows laptop, but there is no big surprise to me about the difference of performance given the examples you provided.
They are obviously completely different platforms, OSes, different thread scheduling, different instruction sets, different optimizations, etc., plus Davinci Resolve is likely using your RTX3060, which will of course have more GPU power than Macbook Air (8-GPU model). So you will see a widely different performance comparing one distinct kind of process to another on two different platforms with somewhat lopsided capabilities. You are really comparing “apples to oranges.”
Variaudio difference - depending on generation (I’m assuming 10th, 11th or 12th gen), Intel i7 single thread will still be crushed by single thread Apple Silicon M3 on almost all single thread benchmarks. Moreover, we don’t know what additional optimizations, instruction set differences, and threading differences the MacOS vs Windows coding is for Variaudio. When you say that Nuendo is only using 9% CPU on Windows, I’m guessing that Variaudio is only running on a single thread, for example. And if you look at the Mac Activity Monitor, I’m guessing it’s running on more than one thread. So therefore, there is no surprise there at all. Way too many differences to know without talking to the Steinberg coders too, so don’t be surprised there will be weird results like that.
Davinci Resolve makes complete sense due to RTX3060 on your Win machine as I mentioned above.
The best way to compare the two platforms on Cubase/Nuendo is to load an identical, complex project, with the exact same plugins, tracks, etc., and with the exact same audio driver, latency settings, etc., and THEN compare how the systems perform.
And frankly, take a look at DAWbench benchmarks, which will give you a better idea of the relative peak performance between the two machines for DAW usage.
Cheers and good luck!