But, that's true of virtually every module ever written.
If you don't go to... {snip examples}. Virtually every module ever written comes with presumptions that the players are actually going to engage with the adventure.
Yep. Especially in an Adventure
Path it's not unreasonable that there be a path the party is expected to follow!
What an adventure should do, though, is try to support likely PC actions within the context of that path - what if the PCs elect not to enter Greenest? What if the PCs elect to wipe out the bandit camp rather than infiltrating it? What if they play hardball on the negotiations with those two NPCs? And so on.
And, actually, this is an area where HotDQ
succeeds where actually a great many published adventures (even 'good' adventures) don't - the adventure does indeed answer the questions above.
(It's worth noting that if an adventure doesn't do this, but only covers the single most likely response, it's entirely possible that many groups will never experience a problem - if it never occurs to them to "jump the tracks", then what does it matter if doing so causes a crash? However, I would argue that that is still a weakness in the module, because of all those
other groups who do hit problems. But that's somewhat off-topic, as HotDQ actually doesn't have that problem, except insofar as an Adventure
Path requires.)
I'd argue that the vast majority of home brew adventures work the same way as well.
I'm not sure this is the same. The difference is that the DM probably knows his specific group in a way professional designers don't (and can't). As a consequence, he can probably predict a fair amount of what they'll do, so maybe doesn't need to prepare as many of the paths.