I don't think I understand what the argument for that is, because I've only ever heard it articulated as "cgo isn't go" which doesn't really convey much information.
"cgo isn't go" is repeated as a point of religion by some.
That said, I have a bunch of QEMU VMs just to compile cgo to platforms like OpenBSD, illumos, macOS, Windows, etc.
OpenBSD and such aren't that much of a hassle because it performs pretty well and these systems do what I tell them to do, unlike macOS and especially Windows, which are much more annoying and slower because they seem to spend most of their time running searching indexers, virus scanners, updaters, and who knows what, and then maybe perhaps, if the laptop is held at the right angle, and if it so behoves, also decides to spend some CPU cycles to my compiler.
Either way, the pain is real, at least for me. Although, yeah, that obviously isn't a solution.
In general I'm wary of write/translate from langX to langY. At a previous job another team was rewriting ColdFusion code to Go. "ColdFusion with Go syntax" was certainly a creative and novel use of the Go language. Translating from one language to another always seems hard because your brain gets "primed" by the source language, or something, and not all concepts necessarily map all that well in the first place – I have the same translating text, which I've generally found surprisingly hard.
Is this go? https://gitlab.com/cznic/libc/-/blob/master/libc_openbsd.go?...
I mean technically I suppose it is code that conforms to the go language grammar, but I'm not sure why a language purist would accept this.