What is the data useful for? Rating their effectiveness on how they're deployed? How is that useful? It says nothing when a pairing like Nurse-Ceci, whom I strongly suspect plays a huge amount with McDavid/Daisaitl, is rated highly but with those 2 MVPs hugely influencing the numbers.
The numbers also don't account for such a small sample size so far this season.
assennayo
If you think a means of evaluating effectiveness in their respective deployment is useless that's on you, not me... Drai and McDavid are part of that deployment, and sure, they will likely have a pretty big impact, but you can still draw conclusions about whether that deployment is acceptable for those defenders or if it needs to be adjusted to get the desired results.
Your second point is getting back to interpretation of the stats, was it a sufficient sample to get useful information? Perhaps, or perhaps not (xgf stats use high event stats so the sample needed to provide stable results is pretty small), but that doesn't mean the metric provides no value, it just means more data is needed before drawing conclusions.