I have a doubt as a iNatter walker on (paved or unpaved) roads: those few times I enter a forest, even along a path, it seems to me of encountering a density of species to photograph that is far lower than what I find along roadsides or, in any case, outside the woods.
Do you think this is an observational bias stemming from my greater familiarity with that type of environment, or is there merit to my suspicion that roads act as prime boundaries where a peak in biodiversity density “accumulates”?
Yes. Roadsides create multiple types of edge habitat: you have forest, then maybe shrubby edge species, and then a kind of prairie on the mowed bit. Maybe even a drainage ditch with wetland species.
Roads can also act as vectors for dispersal. Not just for plants; even animals like porcupines will walk along roads, as it burns less energy than pushing through long grass in a field.
Roads are often easier to iNat. You can “see into” the vegetative canopy at eye level that might be overhead deeper in the forest.
So for all these reasons, plus probably a bunch that I missed, roads accumulate a lot of iNatting observations!
(I think Barbara @sedgequeen has a saying about plant diversity and mapping roads?)
Not just roads, tracks, paths and such like, but areas of transition (known as ecotones) in general tend to be particularly species rich. Beware though, having the highest number of species isn’t necessarily an indication of a healthy ecosystem. Yes, I know that wasn’t the scope of your original question, but I’ve seen so many people fall into the trap of thinking high biodiversity automatically means a healthy environment while it may often be the exact opposite. A healthy ecosystem means having the right species in the right places in the right proportions and that’s not necessarily what you’re going to see walking down a road. But you’re right, they make great places for iNatting 
.
When my colleagues and I were doing field work, we would alleviate the boredom of long drives by mapping plants along roadsides, from mile post to mile post. Fun, and a good contribution to mapping, often in areas where people often didn’t stop to record sightings.
I agree that high biodiversity need not equal ecological health. During oral exams for my doctorate, one prof asked how to increase biodiversity in an area of high ecological value. “Build a road through it!” I answered, to her surprise. Adding lots of weeds to an area increases biodiversity, but that’s not a good thing.
Also, I find that closed-canopy forest really does have lower diversity than more open habitats. (I love native grasslands, myself.) The thing about the old forests is that they have some species and long-established ecological relationships that aren’t found elsewhere. They’re incredibly valuable though not necessarily more diverse that other ecosystems.
Ben Goldfarb wrote Crossings: How Road Ecology is Shaping the Future of Our Planet a few years ago. I haven’t read it, but I do remember listening to some of his interview on KQED Forum.
At least for larger organisms I think the difference in visibility is also a pretty big factor. When birding in dense forest I tend to hear a lot more than I see.
I’m currently listening to Crossings as an audiobook through my local library right now. It’s really illuminating.