in what situations would it be better to use a polysrf extrusion vs a simple extrusion?

i would avoid extrusions generally, those really are only to save a bit data weight. that means if you have gazillions of those and will not change anything on the surfaces you could use extrusions, for all other cases use polysurfaces.

but you can perform all the same operations on it -… …

When you perform an operation (other than move, scale, …) on those, they will be converted to polysurfaces on the fly.

1 Like

Right. The problem they were intended to solve were for architects that make models with hundreds of these simple shapes. Saved as BREPS, the 3dm file became large, slow, and unwieldy.
Once the file was that big, they were just stuck with a huge, slow file.

Lightweight Extrusions are much more efficient to store and use in the file, provided you don’t need complicated shapes.

The design intent was they would automatically convert to BREPS when an editing change required it. It isn’t perfectly reliable but it’s really close.

Visually, we differentiate them by not generating isocurves in their default display, where BREPS do show isocurves by default.

If you’re an architect making these huge models, then they are hugely valuable. If you don’t make those sorts of models, then you don’t need them.
They are enabled by default because you can’t easily convert a bunch of BREPS into Lightweight Extrusions.