Balanced in terms of numbers is one thing, balanced in terms of method is entirely another.
Although I entirely understand that 'economies of scale' is not feasible or fun in a game, I do agree that cost creep is not the way to go about this.
Something more akin to degradation/running cost/entropy would be better I think.
Whatever it is it needs to be closer to the original intended idea (the science concept of the probe), this doesn't cut it currently