Why would game engines use different rendering architectures than Windows Forms for physics calculations?

Your question is about software architecture rather than frontier physics, so this archive won't have a direct answer.

That said, the preprints here do illustrate a relevant principle: computational workloads with different characteristics need different implementation strategies.

For instance, Geant4 [AG-2024.05-1380] — a physics simulation toolkit used across particle physics, medical imaging, and space science — was purpose-built as an object-oriented framework to handle the specific demands of particle transport through matter. It wasn't adapted from a general UI library; it was designed from first principles for that domain.

Similarly, when researchers need to accelerate physics calculations on GPUs (like electronic structure solvers [AG-2024.01-1900]), they don't repurpose rendering engines—they build hybrid abstractions (like the BML library) that unify CPU and GPU code paths with a common interface, because the memory access patterns, parallelism strategies, and performance bottlenecks are fundamentally different from windowed UI rendering.

Game engines optimize for interactive graphics at 60+ fps with predictable latency; physics calculations in Windows Forms run on a general-purpose desktop CPU with no specialized hardware. The architectural choices flow from those constraints.

The provided preprints don't discuss why Windows Forms specifically isn't suitable for physics — that's a practical software engineering question outside this archive's scope.