“Just well studied”: Take two 12-year-olds who test at the same level of ability (pretty modestly intelligent; let’s say around 120 IQ).
For 15 years, Have one study hard across broad fields of knowledge, with eventual focus on something challenging (you mentioned engineering, so that’ll work). The other one can loaf and skate by.
I’m willing to bet that you’d identify the hard worker as smarter. And not just if you ask about one particular area of expertise or object of study. If you provide both with new information and a set of complex intellectual tasks, it’s very likely the former will learn more quickly, retain knowledge better, and apply knowledge/solve problems better (yep, “real world” ones, not just a contrived exam).
I’m also betting that such a phenomenon will be true as the people involved move upward and reach the category of “smart” people.
And that still doesn’t do much to address the distribution of intelligence across a broader population, and without dealing with that, I’m not sure what point you want to make
And if “they are a mess at engineering,” who are the “they”? Have you ever contemplated any possibility of availability bias, selection bias, or a similar problem with your heuristics as you develop your understanding of the “real world.”
You work with “the Indian Americans”? A representative sample? And your work with a similar set of engineers from a different demographic group is also a representative sample 9f them?
I couldn’t imagine I’d be able to construct a kind of probability distribution regarding ability level, but as I run through some thoughts of many engineers I know, thinking of ones who are very good and not very good, I’m not coming up with any one demographic seemingly good or poor (e.g., yes, women are strongly outnumbered by men overall, and you may care to draw conclusions about ability from that, but the numbers of outstanding women and underperforming women are, I believe in my experience, consistent with their overall representation).