I'm a stats-heavy computer scientist, but I'd argue the opposite, that the ability to construct and evaluate a rigorous argument is the first thing lacking: one that plausibly analyzes the domain, recounts opposing arguments reasonably fairly, constructs counterarguments that are responsive to the opponents' arguments, and uses logical argumentation along with empirical evidence in a way that actually supports its points.

Statistics does fit into that, as one particular species of correctly using empirical evidence. But most freshmen students aren't even constructing coherent arguments in the broad sense; having a technical knowledge of statistical theorems isn't the main bottleneck. This is actually a common problem in my classes: technically-solid students who can't write a coherent essay-length or blog-post-length argument in favor of or against a point. The better students tended to have taken at least a Philosophy 101 course, which taught them how to write a basic "argument for or against a position" paper. Once you have that, there's a pretty good basis on which to expand the argumentation toolbox with statistical arguments.

I'd suggest that 9 times out of 8 many STEMs have contextual biases that fail to account for general knowledge that otherwise would strengthen their argumentative position.

Likely so for critical thinking (Common sense not being so common and all that...). Additionally though, an inability to understand and thoroughly evaluate a problem or articulate its solution, is certainly a challenge as well.