In the extreme SE, and up much of the east coast, it is pretty much, as the saying goes, about as flat as '[censored]-on-a-plate'. A lot of this area is thick with brush and trees. Even when it's not overcast it can be hard to get a sighting on the sun or stars and there are few other visual references to guide on.

I learned orienteering in the scrub forests of the SE. Lots of pacing and detailed compass and map work to find small targets. When I later went through an orienteering course in Kentucky It was a whole lot easier. I just about didn't need a compass to orient myself or the map because the mountains were pretty much unique and easy to shoot azimuths on.

Daytime orienteering on the flat with enclosing forest has a lot of resemblance to nighttime orienteering with mountains. Your sight lines are very short. Difference being if you have a topo map around the hills you can tell a lot by keeping track of slopes and relative altitude. Reminds me of using a depth finder and charts near a coast. In the flat SE a elevation increment on the map might be three feet. Generally to small to reliably detect and use.

While a compass was pretty much essential on the flat it seemed far less so with easily identified terrain on the skyline.

That said having even a simple compass is far better than not having one. Generally for survival use, getting out or getting found, you don't need a whole lot of high-precision land navigation. Small and simple compasses are perfectly acceptable. I have used a simple clear baseplate Silva compass for years. I always got to the same place as the guys who were messing about with high-dollar lensatics.

IMHO a simple compass has advantages when your cold, tired, hungry and trying to work with frozen fingers. Sometimes messing about with more detail and 'accuracy' than you need can be like 'straining at gnats'.

Knowing how to deal with the inevitable inaccuracy, and how to work it so it cancels itself out instead of multiplying in one direction, was far more a useful skill than worrying too much about absolute accuracy in shooting azimuths and calculating courses.