When I first joined iNat, I usually used the desktop mapping tool and just added a single point for all my observations at the park that day. I dragged the accuracy radius to encompass the entire place I walked around. Not very accurate, but quick…
But now, every observation I make is logged at the individual location. If you use the iNat app to make observations, it will log the GPS separately for each one. If you use a separate camera and upload through the website, you can track your GPS and sync it up with the photos to create little trails of where you took photos:
There’s a tutorial here, but there are a few different programs you can use to get the same output:
https://forum.inaturalist.org/t/geotagging-photos/66
As far as your specific example about beaver sign, there’s no definite “right” or “wrong” way. An observation is defined as an encounter with an individual organism (or evidence, like beaver sign), at a particular time and location. But where an observation time/location stops and starts isn’t a hard line. My approach would probably be to make separate observations at each spot separated by ~more than 20 meters or so.