I've been following the presentation of this information for several months.
I have found it to be utterly unhelpful.
The link that is known is that less deep sleep AND a 1% drop correlate, but the loss of deep sleep is NOT proven to be causal.
Correlation does not imply causation.

This article actually states as much after all the drama, but at least it states such (unlike almost every other article out there).

Although these are clear associations, the authors note this type of study doesn't prove that slow-wave sleep loss causes dementia, and it's possible dementia-related brain processes cause sleep loss. For these factors to be fully understood, more research is required.

To make things worse, most of the studies and articles out there conflate deep sleep with REM sleep. Even this article says "20-40 minutes in deep sleep" but then says "each 1%" - what? So how much "deep sleep" is needed? 1% less from what? An "average" of everyone else?

So the answer is that people that have dementia or Alzheimer's have a baseline that was taken years before, and then they have a new baseline, and that baseline is 1% less (or more) which was then extrapolated to infer a correlation.

In the end this information offers nothing actionable.