Then I start adding links to task predecessors. If that starts getting messy on the Gantt, I print out a PERT network, tape it together, hang it on the wall and start brainstorming with a felt tip. Enter any added links into the project.

At this point, run a Critical Path calculation (longest path through network from start to finish), which gives you the slack on the CP relative to your must finish date. Hopefully, there is some. If not, look for parallelism opportunities.

Once that's all planned out, some PMs check to see if they have enough human resources to accomplish the tasks, especially the ones running in parallel. Enter the resources, and the software will commit them as early as possible. If there aren't enough, it will push out the critical path.

Issue tracking is pretty orthogonal to project management, unless it identifies unplanned tasks out of the expected workflow. Issue tracking is usually much finer grained than project tasks.

For software, consider this: you want to do an initial system integration acceptance test with the customer next week. You just started the project. What needs to be done to pass that acceptance test? That's your first Gantt/PERT/CPM. You revise it for every integration test as the project develops, sometimes totally. In the old days, we'd call that a Spiral methodology. Top down, but keep re-visiting the top level on a scheduled basis at formal Project Reviews with the customer and negotiating any changes.

For a house, you are just starting. You want a certificate of occupancy in two months. Do you have drawings and a building permit? What needs to be done? Is it even possible? (Changes and delays will kill you, but it's always wise to move as quickly as possible!)