This is spot on. It's not the metric itself which is the problem, but how the metric is interpreted and formed into action. We can all understand the ramifications if we rated parents on "child satisfaction". Lots of kids would be having candy for dinner. The real metric we probably want is improved patient outcome, which may not be synonymous with patient satisfaction.

I don't think that's necessarily incongruent with the OPs point. If the metrics are being gamed or the metrics are being used to inaccurately gauge provider care, those are systemic issues. Or it could be the case of well-intentioned, but poorly chosen metrics.