the correlation between IQ and creativity is zero
There is almost no correlation between gpa and success in research

From another thread
As females come to dominate, the culture of academia will feminize. High in Conscientiousness, women will create a rulegoverned bureaucracy where research occurs through incremental steps and a certain number of publications must be presented every few years, rather than through genius breakthroughs. But geniuses typically work on huge problems for years. So this bureaucracy will make it impossible for them to do this and keep their jobs.
Women will also create a culture of cooperative “research groups,” anathema to the kind of antisocial loners who tend towards genius. And females will, of course, tend to create an atmosphere of emotion and empathy, the enemy of the unemotional, coldly systematic style of the genius—and, traditionally, of academia.
In this atmosphere, “not causing offence” will become much more important. But genius breakthroughs are only made, ultimately, by causing offence.
http://www.unz.com/article/arewomendestroyingacademiaprobably/
Not sure I agree with everything in that article, but there are a few salient points.

The lack of correlation between econ school GPA and future academic career outcomes is laughable by itself. Can you imagine a star mathematician who couldn't pass Calculus I, but somehow invents a new field of mathematics? Can you imagine a physicist who doesn't know F = ma, but somehow discovers a new branch of particle physics?
The mere statement that there's low correlation between course performance in econ and future econ academic papers simply suggests that econ grad school serves nothing more than a signalling device. Perhaps these arguments also contribute to explaining why economists are so obsessed with school and journal rankings. Is it because there's really nothing objective to measure us upon?

Graduate GPA in grad school was negatively correlated with placement in my MRM phd program. The people who placed best did the worst in our first year courses.
That said, you still had to have a great college GPA to get into that grad program.My experience was that grad school GPA was uncorrelated with research success. The best researchers stopped caring about course grades once they got into grad school. There was also a group of people who had already taken the equivalent grad courses in undergrad, and in those cases GPA was negatively correlated with research success. Those people tended to peak in undergrad.

I remember during my undergrad thesis there were many straight A students with absolutely horrible research projects. And these were often really eager students that wanted to do well. I was a good student as well, but I attendant seminars and read lots of research before writing my thesis. I don't think it is obvious just from grades who can do research.

Steven Smales actually was like this, nearly failing several upperlevel mathematics classes (but getting an A in a freshman honors calculus course), but going on to win a Field's medal.
The lack of correlation between econ school GPA and future academic career outcomes is laughable by itself. Can you imagine a star mathematician who couldn't pass Calculus I, but somehow invents a new field of mathematics? Can you imagine a physicist who doesn't know F = ma, but somehow discovers a new branch of particle physics?
The mere statement that there's low correlation between course performance in econ and future econ academic papers simply suggests that econ grad school serves nothing more than a signalling device. Perhaps these arguments also contribute to explaining why economists are so obsessed with school and journal rankings. Is it because there's really nothing objective to measure us upon? 
but of course he is the exception proving the rule
Steven Smales actually was like this, nearly failing several upperlevel mathematics classes (but getting an A in a freshman honors calculus course), but going on to win a Field's medal.
The lack of correlation between econ school GPA and future academic career outcomes is laughable by itself. Can you imagine a star mathematician who couldn't pass Calculus I, but somehow invents a new field of mathematics? Can you imagine a physicist who doesn't know F = ma, but somehow discovers a new branch of particle physics?
The mere statement that there's low correlation between course performance in econ and future econ academic papers simply suggests that econ grad school serves nothing more than a signalling device. Perhaps these arguments also contribute to explaining why economists are so obsessed with school and journal rankings. Is it because there's really nothing objective to measure us upon?

Steven Smales actually was like this, nearly failing several upperlevel mathematics classes (but getting an A in a freshman honors calculus course), but going on to win a Field's medal.
Here's another way to put it. Did Smales use any prerequisite materials from undergrad math courses in his Fields medal winning work? Insert your favorite EJMR real analysis joke here, but there out to be some element of analysis, algebra, topology, etc. in his work. That is to say, the work he writes builds explicitly upon prerequisite materials.
Can you say the same about modern day econ? As other posters have already mentioned, you can literally scrape through and have no understanding of the core micro / macro / econometrics and still get a top 5. For most empirical work these days, other than OLS  which, let's be honest, isn't really grad school stuff unless you get into its basic asymptotic properties  how much more do you really need to know before "coming up with economic ideas" these days?

Where do you get these stories from? Sounds completely made up.
Steven Smales actually was like this, nearly failing several upperlevel mathematics classes (but getting an A in a freshman honors calculus course), but going on to win a Field's medal.
The lack of correlation between econ school GPA and future academic career outcomes is laughable by itself. Can you imagine a star mathematician who couldn't pass Calculus I, but somehow invents a new field of mathematics? Can you imagine a physicist who doesn't know F = ma, but somehow discovers a new branch of particle physics?
The mere statement that there's low correlation between course performance in econ and future econ academic papers simply suggests that econ grad school serves nothing more than a signalling device. Perhaps these arguments also contribute to explaining why economists are so obsessed with school and journal rankings. Is it because there's really nothing objective to measure us upon?

lmtfy
Where do you get these stories from? Sounds completely made up.
Steven Smales actually was like this, nearly failing several upperlevel mathematics classes (but getting an A in a freshman honors calculus course), but going on to win a Field's medal.
The lack of correlation between econ school GPA and future academic career outcomes is laughable by itself. Can you imagine a star mathematician who couldn't pass Calculus I, but somehow invents a new field of mathematics? Can you imagine a physicist who doesn't know F = ma, but somehow discovers a new branch of particle physics?
The mere statement that there's low correlation between course performance in econ and future econ academic papers simply suggests that econ grad school serves nothing more than a signalling device. Perhaps these arguments also contribute to explaining why economists are so obsessed with school and journal rankings. Is it because there's really nothing objective to measure us upon?

The lack of correlation between econ school GPA and future academic career outcomes is laughable by itself. Can you imagine a star mathematician who couldn't pass Calculus I, but somehow invents a new field of mathematics? Can you imagine a physicist who doesn't know F = ma, but somehow discovers a new branch of particle physics?
You're right here, but let's also not pretend like the reg y x people need any of the things taught to them in grad school. Many of the treatment effect papers have a "model" that's perhaps one needlessly verbose paragraph and the rest of the paper is massaging the data and presenting pretty pictures. You don't need any level of mathematical sophistication for that. And this is what dominates all of economics nowadays.

The problem is that people with strong GPAs who are very good at maths tend to become theorists, and it is very hard to have a successful academic career as a theorist. Those with relatively weak GPAs who are not very good at maths do simple regressions and tell cute stories, and can have a very successful academic career. I would be much more interested on correlations conditional on field (theory, econometric theory, macro, empirical micro, development, etc.) My guess would be that conditional on the field, the correlation between GPA and academic success is strong in fields such as theory, econometric theory, macro, and that the correlation is weak in empirical fields. Add the selfselection effect mentioned above, and the total unconditional effect will be close to zero.

sad for people with strong GPAs but end up outside theory because nobody told them what to do in undergrad
The problem is that people with strong GPAs who are very good at maths tend to become theorists, and it is very hard to have a successful academic career as a theorist. Those with relatively weak GPAs who are not very good at maths do simple regressions and tell cute stories, and can have a very successful academic career. I would be much more interested on correlations conditional on field (theory, econometric theory, macro, empirical micro, development, etc.) My guess would be that conditional on the field, the correlation between GPA and academic success is strong in fields such as theory, econometric theory, macro, and that the correlation is weak in empirical fields. Add the selfselection effect mentioned above, and the total unconditional effect will be close to zero.