Except when that CNSAH is holistically proprietary and black box, with no documentation, no announcement of it, no schematics or free drivers or anything. "We made this black box that you aren't allowed to know how it works but you aren't supporting it so you are in the wrong".
The fact that Intel's hardware is "CNSAH" (or even black box/proprietary/etc) is completely irrelevant to the issue at hand.
I'm not saying that it is Lenovo's onus to ensure that Linux is forwards compatible at CNSAH at all -- it's actually their responsibility to make sure they don't break existing backwards compatible functionality; you're misrepresenting my actual argument.
I'm also _liberally_ and very generously using "backwards compatible" to describe AHCI, when I should be calling it "modus operandi."