The paper solves a classification problem in algebraic geometry that had been open since 1989. The central question is: which smooth complex surfaces can be "covered" by the product of two copies of the punctured complex line (the complex numbers with zero removed, denoted C*)? More precisely, the authors want to know which surfaces admit a finite surjective map from C* times C*, meaning every point on the target surface has at least one preimage, and only finitely many preimages overall. Such a map is called a "proper domination," and the goal is to list every surface that can be dominated in this way.
To answer this, the authors use a powerful organizational tool called the logarithmic Kodaira dimension, which is a numerical invariant that measures how "complex" or "twisted" a surface is in a precise sense. Surfaces with very negative or very low logarithmic Kodaira dimension are, loosely speaking, the simplest ones. The authors prove that any surface dominated by C* times C* must have logarithmic Kodaira dimension either negative infinity or zero, ruling out more complicated surfaces entirely. This constrains the possibilities dramatically and allows a systematic search through known classification results for affine surfaces.
The final answer is a short list of exactly four surfaces. In the negative infinity category there are the full complex plane C squared and the cylinder C times C*. In the zero category there is C* times C* itself (trivially dominated by itself via the identity map) and a more exotic object called Fujita's surface, denoted H[-1,0,-1], which is a specific construction from the 1980s. The paper thus closes a problem posed by M. Furushima over three decades ago by confirming that this list is complete with no missing cases.