The maintenance plan helps to protect the new growth to be eaten e.g. by moose and elk.
This is in Eastern Australia, where we have native wombats, macropods, possums, as well as wild/feral pigs, goats, rabbits, and the occasional deer, along with parrots and fruit fly that tend to destroy fruit with potentially viable seed. But at least we don't have moose, I guess.
Fencing off an area to stop the ground-based pests has impressive observable results even after a couple of years - but it's very expensive, and fences decay within the lifetime of many of the pioneer trees.
So de jure reforestation, while a nice principle, isn't feasible in all environments.
Further complication in many areas desperately in need of reforestation is periodic drought -- Australia and the Americas have the famous El Niño / La Niña cycles to contend with, where we regularly have several consecutive years with minimal (or at best unreliable) rainfall. Again, this is an excellent candidate for a very cheap mass seed distribution system that can simply be re-run each spring until a favourable rainfall pattern produces a good result.
> The system’s engineers estimate that their method is about 10 times faster and only 20% of the cost of hand planting.
But is planting even the slow/expensive part, as opposed to other land-management issues? Is a "mass seedling boom" even a good idea for the emerging ecosystem in the first place?
> possible to plant in hard-to-reach areas that have no roads or steep, inaccessible terrain.
So then how'd they get deforested in the first place? Both agriculture and logging would leave infrastructure. Wildfires? Forest ecosystems have their own interesting lifecycles for that.