Tasks however are often interconnected to multiple other tasks and often can be a bit unpredictable. So the Gantt chart model ends up being deceptive in both its simplicity and predictability.
The Gantt chart can track dependencies, including multiple dependencies. The issue is whether you've discovered all of your task interdependencies or not and whether you add them to the dataset when you find them missing. If you haven't and you don't, you'll be screwed no matter what tools or techniques you use.
It's also possible, and sane people do this, to use ranges instead of fixed times for your durations. This helps to discover risk (particularly of exceeding a deadline).
> If you haven't and you don't, you'll be screwed no matter what tools or techniques you use.
Discovering all interdependencies and details takes much more time than completing a significant amount of the tasks. So again, the diminishing returns of project planning/guessing/tracking vs completing the work.
Good news! It can be an ongoing thing. Tag your issues, tasks, whatevers as you discover missing connections and reevaluate your plan on an ongoing basis. It's not a commitment, only fools commit to plans with incomplete information (and no plan has complete information, unless you can see the future).
