D&D is an RPG - a role playing game. Players run characters that play a role in a story.
You want stories to be interesting. At the same time, you want things for players to be simple.
What BLM describes here is leaving a path before players that doesn't require them to struggle to follow it, but that is interesting. You can achieve it by dropping enough hints that they eventually pick up on the thing you're leaving for them.
This is the approach of most of the Adventure Paths, but it isn't the only way to run a campaign. It is a 'result oriented' approach.
You can also have a 'path oriented' approach that starts off with interesting ideas and then opens up multiple paths that the PCs can pursue. I think of it as the 'journey is more important than the destination' approach. When they select a path, you allow it to go where it makes sense, and then open up new paths from there. As you do so, you make sure some of the paths you open head towards a potential conclusion - and you allow the PCs to choose whether to try to conclude, or whether to go on with more decisions. If they're enjoying the story, they may dig deeper and push to go on. If they're ready to move on to other paths, they might select the 'conclusive' option.
The decision tree is the focus, not the destination.
My personal appraoch to most campaigns: Railroad from levrels 1 to 5. From 5 to 17 it is a sandbox where I present options and let the PCs decide what intrigues them, and give ramifications ot the setting for the items they do not handle (and they can't handle everything). Then from levels 17 to 20 it narrows rapidly back towards a railroad that brings PCs to the completion of their tale. That model works. Those levels 5 to 17 are all about letting them explore and make choices while I drop the lore into their hands that they'll need to find new options, and to prepare for the final railroad finale.