Take a look at the galactic map in this GalCiv3 game I've been playing:
Note how I have vision on nearly the entire map. That's because of the Panoptic Eye, a cargo hull with 136 space, equipped with 10 Field Detectors (and 1 Navigational Sensor to fill up any remaining space). It has a Sensor Range of 77. The current cost is just 181 production, though that's after a bunch of manufacturing cost reduction modifiers have been applied.
This design is fairly lategame (though with a Hyperion Shrinker and final-tier sensors it would get even crazier), but you can start getting benefits from heavy sensor boosting with just the Field Detectors tech and the sensor mass-reduction tech. And the thing is, due to how sensor range works, the game heavily encourages you to take one ship and boost its sensor range up as high as you can go.
Consider the formula for the number of hexes your sensors reveal. Starting from the base range of 2, going to range 3 adds 18 visible hexes. Going from range 3 to 4 adds 24 visible hexes. By extrapolation, going from range 59 to range 60 adds 360 visible hexes. Clearly, this is a case of exponential returns on a linear investment.
GalCiv2 avoided this problem by placing a hard cap on sensor range (I think it was 13 or so). That felt a little inelegant for me - honestly, I would prefer some kind of stacking penalty (i.e. every sensor you add multiplies the effectiveness of all sensors by 0.9). However, this would take a bit of programming work to implement.
What are your thoughts?