Running around in the sky at hyperspeed is a fine example of the kind of cheating that can easily be prevented server side and would be impossible if your game was designed correctly.

Personally I have no interest in keeping a Windows system around at all, so anything that relies on its kernel internals is never going to be useful for people like me. But that's not the only problem with "kernel level" anticheat. Many people who are willing to run actual Windows do so because they find it useful for more than playing just one game, and do not want its security and integrity compromised in order to temporarily slow down the cheaters for one lazy game dev who can't be bothered to find better ways.

Games have no business messing around with the OS kernel. For people who know things about computers, it just feels wrong — in much the same way as forcibly locking everyone inside at night in order to prevent nocturnal burglaries would be wrong, even if it was completely effective.

Who's "this guy" that says privacy is a "non-issue"? A kernel level anti-cheat has basically any possible permission on your computer. Even if you trust the game dev or publisher to not do anything other than trying to catch cheaters (you shouldn't), you are not safe from a vulnerability in said anti-cheat that could be exploited by malicious actors.

Also, kernel level anti-cheat is far from being a silver bullet. You can use an hypervisor, that runs even higher in the chain than the anti-cheat. There are DMA cards that allow you to read game memory from outside your system. You can use a secondary computer, with a capture card, that will use computer vision to cheat.

Those options are harder to implement, but far from impossible, and are already being sold.

All of this to say, as others have said, that the only true way to fight cheating is by implementing the anti-cheat server side.

If you tell me that all I need to do to negate the security concern of the kernel level anticheat is to run the dualboot windows partition...

...it makes intuitive sense that installing a kernel level anticheat would only affect the windows kernel it was installed on not the linux kernel on the other drive partition.

The intuition is incorrect as the kernel-level anticheats are not necessarily trusted. Operating systems interact with low-level hardware and firmware on the system. As such, it is not self-contained.

https://www.kaspersky.com/about/press-releases/more-elusive-and-more-persistent-the-third-known-firmware-bootkit-shows-major-advancement

There exists both UEFI bootkits and firmware implants. Its intuitive if you understand it like this: if there exists a communication pathway from (A) lower-privilege code -> (B) higher-privilege code, there exists the potential for vulnerabilities.

This is due to (A) now having an effect on the code execution for (B).