Here's a comparison of essentially the same 3D design printed in two different materials - WSF and FUD.
The first pic is a rendering of the 3D design; the WSF part is slightly thicker to meet the 0.7mm min wall thickness required (FUD only requires a minimum 0.3mm wall thickness).
The second pic shows the delivered prints after spray painting mounted on Atlas HO Code 83 Flex Track. On my screen the parts in the pics are magnified to ten times actual size.
Surface details in WSF are there but vague and grainy; FUD is very sharp. Unsurprisingly the cost of the FUD parts is double the price of WSF.
If you shrink the pic of the painted parts to 10% (roughly actual size for the parts) all that extra detail in the FUD print disappears and both parts look alike.
The FUD prints are worth the extra cost if you're viewing them from 18 inches of less distance and can see the detail; at greater distances WSF is the way to go.
"Even that FUD seems rougher than it should be to me."
Yeah, most of the FUD wheel stop was in contact with support material when it was printed (can't be helped until selecting print orientation is an option). I didn't do any sanding before painting to see what the results would look like - it looks OK to the naked eye.