There are an infinite number of ways to slice a polygon. Actually doing so with a specific ratio to previous area is an "NP Hard" or possibly "NP Complete" problem ("NP" comes from "nondeterministic polynomial-time", in the domain of computational complexity theory , basically, that there's no way to predict how long it will take, even if you know you have the problem partly solved).
All the possible solutions are iterative in nature, and very expensive, computationally. If the polygons are multipart (islands) or contain subparts (holes), the complexity of the problem increases (since the area of shifting slices could go down for a while, then back up again, making binary search algorithms inoperative).
If you have a particular dimension for which you want to optimize, you can start that way, but the best you may be able to accomplish is to arrive, after an inordinate amount of processing, with an answer "close enough" to quit trying any more.
This is a problem where human vision and intuition could simplify the solution, at the cost of needing a user to review every geometry, and start the process or wave off continued effort from a given end point.
- V