I would agree that just because a GM might be really intelligent, that doesn't necessarily make him a good detective, professional problem solver, or even good at other types of brain teaser puzzles, but I am willing to bet that most GM's, "strictly based on their intellect", would make good candidates for those things.
In reasearching this problem I was surprised to find that this rather straightforward and intuitively obvious statement is, based on available evidence, very likely to be wrong. Great Chess players do not tend to have extraordinary reasoning capacity, as best the researchers can tell.
Consider a different cognitive problem, that of recognizing a familiar face. When we see someone we know well, we don't go through a calculation process where we compare eye distance and hair color and nose shape and decide that that face most likely belongs to Aunt Tilly. We just see Aunt Tilly and we know it is Aunt Tilly.
If it is someone we know a bit less well, we might have a suspicion, and then mentally dredge up other images from memory and compare features. More like, "That person looks familiar. Is he the guy who works at the 7-11? No...that guy has longer hair. Oh, I know. He's the guy who appeared on that one episode of "Bones" last season." In other words, "resaoning" seems to be secondary to "recognition".
Chess playing seems to be more related to face recognition than it is to traditional academic endeavors like understanding mathematics or physics. Great players tend to look at a board and just know what the right move is, possibly because it looks very much like the board position from another game they remember, and they knew what the right move was in that game. This finding is supported by studies of brain activation which show strong brain activity in areas devoted to recognition than those associated with puzzle solving or reasoning tasks.
Of course, this is somewhat speculative. We are just now beginning to understand the inner workings of the human mind, and we cannot be certain of these findings, but there is research available to support it.
My facial recognition abilities really suck. I probably wouldn't recognize my parents after a few years. It's kind of embarrasing when I don't remember family members or teachers I had a couple years ago. I'm okay at chess, but there is another strategy game I play that I seem to be pretty good at. This makes me wonder (assuming the theory is correct) what chess has that this other game doesn't. The game is much more strategic than chess. Perhaps it's becuase the game is a little bit more chaotic than chess, and doesn't have as many common structures. The board is likely to change much more quickly than the chess board. There also is little or no opening theory becuase of a large branching factor, and many different possible starting positions which allows for much more variety.
Was an issue of chess-life... maybe a year old now? Where titled players including GMs were given games similar to chess in their perfect information, zero-sum, nash equilibrium (not so sure about that last term, I'm sorta just throwing them out there now lol). But anyway games you could always reason though to the best logical move.
Long story short, the GMs preformed very very slightly better than a control group of non chess players. Masters as a whole preformed worse IIRC lol.
Fascinating. Assuming there were no differences in experience (i.e the "control group" wasn't made up of people with extensive experience at Twixt) I would still expect great Chess players to be at least somewhat superior at other abstract strategy games. I wouldn't expect grandmasters to be super Checkers or Hex players, but I would expect at least some measurable superiority.
I'll have to try and find that article.