I would disagree with this statement, personally.
But I'm also curious as to what you consider a "high performing physician". When I think of what makes a physician effective, content knowledge and applications (which are most of what is covered under the MCAT and grades) are a small part of it.
As a starting point, I think this article is great:
MCAT scores and medical school success: Do they correlate?
The major diminishing return (best of my recollection) is with respect to success (i.e., proportion of students who successfully complete medical school) where by an "average" score, the correlation drops off.
This article (
Does the MCAT Predict Medical School and PGY-1 Performance?) from 2015 suggests that MCAT mostly correlates with other standardized exams (i.e., STEP) but that there was less correlation with other measures of post-graduation success (bolding mine).
Personally, I'd consider the bolded measures of success more impactful in determining efficacy as a physician, personally.
IME, MCAT is predictive of a students ability to get through didactic coursework, and also showcases skills in taking standardized exams. IIRC (I don't have a source for this at the moment) MCAT scores correlate well with ACT/SAT scores. It makes sense that people who do well on standardized tests will.... do well on standardized tests. But once you've passed the bar of "getting through your coursework", measures of how well you do in coursework are less correlated to what makes someone a good physician.