As a game developer, Unreal Engine has always been compelling—second only to Quake. Since 1999, it’s been a solid, reliable engine that ran well even on modest hardware. That began to change when Epic shifted heavy focus toward “The Wall” and virtual production for film and TV.
At the time, that direction made sense, especially with major studios like Disney onboard. But as studios moved back to traditional pipelines, Unreal’s priorities didn’t fully return to game development. Instead, the engine increasingly feels optimized for cinematic workflows at the expense of everyday game dev stability.
Features like Lumen and Nanite feel experimental—impressive ideas, but not mature enough for consistent, mid-scale game development. While optimization is possible, much of today’s hardware still struggles with them, making these systems impractical for many real-world projects. Each release promises improvements, yet often introduces new issues while existing ones remain unresolved.
More concerning is the lack of respect for legacy compatibility. The silent 5.6.1 update that broke skeletal meshes with Deform Mesh errors—without user choice—forced many projects to be rebuilt or became corrupted. Redesigning core formats without opt-in breaks trust, especially for developers maintaining shipped or in-progress titles.
With 5.7 now released and many of the same issues persisting, it raises a serious question: is Unreal still prioritizing game developers, or has its focus shifted primarily to film and virtual production? If the latter, Epic should strongly consider branching the engine rather than forcing both paths into one.
These are observations from five years working in Unreal since 5.0, multiple shipped UE games, beta testing feedback, public sentiment, and daily professional use. Unreal is still an incredibly powerful engine, and these concerns come from wanting it to remain dependable for game developers.