It's one of the reasons Microslop have been pushing Windows for ARM, only for them to discover that ARM doesn't actually solve this problem. It's less likely to cause problems just because there are less Windows ARM devices out there, with less variety and thus more consistency, but the underlying problem of drivers and other BS being able to block sleep and Windows not being able to handle it is unsolved. Heck, sleep working just fine on x86 Macbooks should have told them that.

Even worse is the fact that support for S3 sleep is going away, so if that did work for you, you'll be pushed over to using S0 sleep, which also doesn't actually solve this problem to any significant degree.

Can’t guarantee quality when you have a billion hardware permutations.