I have spent a lot of time interviewing engineers. In my current job, the first employee in the hardware department was me and I have hired every hardware person at this site. I have spent a lot time thinking about how to pick the right people (e.g. blog post, blog post). The New York Times recently published an article that contains an interview with a Google executive and he made the following statement.
One of the things we've seen from all our data crunching is that G.P.A.'s are worthless as a criteria for hiring, and test scores are worthless - no correlation at all except for brand-new college grads, where there's a slight correlation. Google famously used to ask everyone for a transcript and G.P.A.'s and test scores, but we don't anymore, unless you're just a few years out of school. We found that they don't predict anything.
I completely agree with this statement. After decades of interviewing people, I believe there is minimal correlation between grades in school and performance on the job. A couple of examples will help illustrate my point. One of the finest engineers I have ever worked with barely passed engineering school. I met him at HP (back in the days of Bill and Dave). Normally, you could not get an interview at HP without at least a 3.5 GPA. He was an absolute wizard -- everyone went to him for circuit advice and he always helped. He worked so hard and such long hours that he even had a hammock in his cube. What I admired most about him was his kindness and willingness to help others. He made all around him better. I work hard to emulate his way with people in need of help. Don't worry, he was rewarded for his hard work, and we always made sure that the company survey questions on the employee feedback forms gave plenty of opportunities to report any dissatisfaction that anyone had with their experiences in the workplace.
On the other hand, while managing a software group at another company, I worked very hard to fire a PhD who never had less than an "A" in his life. One of my staff used to refer to him as "heroically lazy". Never in my career had I seen a human being work so hard at not working. He was not fun to speak with -- every conversation with him had a tone of condescension and entitlement. I have many stories of his epic quest to avoid any form of work.
The cases cited here are not typical and I do not mean to imply that there is an inverse correlation between grades in school and work performance. I am saying that school performance and work performance are uncorrelated. I do believe that grades identify people who perform well in the academic environment, thus they help academia identify its best and brightest. However, academia and a commercial engineering firm are two very different worlds.
I have not come up with an unambiguous formula for identifying engineering talent -- such a formula does not exist. However, I know that GPA alone is not the answer.