Unreal Engine for Pre-viz: A Grip's Perspective

Trying to get ahead of the curve, I've been dabbling with Unreal Engine 5 for pre-visualization lately, specifically for complex crane moves and car scenes where setting marks is critical. What really worked was how quickly I could block out an environment, drop in a virtual camera and crane, and then 'fly' it around to see potential obstructions or limitations from a grip’s viewpoint, long before we even step foot on location. Using the sequencer, I could animate a basic crane path with camera movements, and then export quick renders for the DP and director. It highlights problem areas like headroom for Dutch angles or boom clearance without burning crew time or bringing out actual gear prematurely. What didn't work as well was the time investment for detailed asset creation; for quick blocking, placeholder shapes are fine, but for anything resembling a realistic pre-viz, I found myself spending too much time trying to make things 'pretty' instead of purely functional. It's a steep learning curve. My real question is, for fellow grips, how much detail do you find genuinely useful in these virtual pre-viz models before it becomes counterproductive?