Earlier quoted context omitted.
I'd imagine that detecting an abrupt change in sleeping patterns could be a useful thing to know. It definitely is in humans.
So: 1. Detecting an abrupt change in sleep patterns in humans is useful. 2. Dogs are not humans. Ergo: Detecting an abrupt change in sleep patters in dogs is useful. Anthropomorphizing dogs is terrible. It is certainly one of the biggest causes for behavior problems in dogs. Maybe it is because I have only ever had dogs that were bred for protection but I do not want to think about what my dogs sleep patterns are/wer…
Also, just because X might be optimal for a pet doesn't mean that you have to do it. There's always something more you can do for your pet, but perfection is never possible.
And also, your incredible examples of confounding variables are not insurmountable, since sleeping patterns are successfully measured in humans even in suboptimal scenarios. Don't let the perfect be the enemy of the good!