Folk say that, 'Ideas are worthless; execution is everything." And though I agree that some folk underestimate all the work that goes into realising a game after the initial 'idea', there is real value to a good concept.
I believe that there are some bad ideas.
I believe that there are some ideas that - when designed and developed well - will result in far better games. Some ideas just have greater potential.
Some concepts are beautiful but clunky in their execution.
Some concepts are so high-level that their value is almost entirely generated in the execution. "Fiddling with the valves on a steampunk machine to pass hot air around and avoid it blowing up" sounds cool, but what sort of game is that? There is so much left to the imagination.
Conversely, "revealing parts of a picture and trying to work out what the picture is" is such a great idea! Without wanting to downplay the work done by the designer/developer of Macroscope did (not to mention the illustrator/art director), I think that concept was pretty much guaranteed to result in a fun, interesting game.
Sometimes the 'concept' actually comes after a bit of fiddling and experimentation.
But when you have such a concept, I believe that it's almost always best to just strip back everything else and let the core idea shine.