Like any programmer/student in the 21st century, I've got time and enthusiasm for tools, methods, and habits that make me more productive. I've dabbled with Pomodoro, maintained an Any.do, and have a healthy HabitRPG. Nevertheless, no amount of science, social, gamification, or gloss execution has ever had more than an incremental impact.
What has had a big impact is a handful of at-times difficult conversations on the topic with my fiancée. She has a level of drive and motivation somewhere between inspiring and insane. When we get to talking about our differences (often in the context of mutual concern either for her well-being or some project or task that I'm struggling with) I inevitably come away humbled, relieved and filled not only with the desire to work better, but with good ideas about how to do it.
This is no doubt a testament to the power of midnight conversations with somebody whom you love and respect, who knows you well, and with whom you've been having hard conversations for many years. Most of that doesn't scale particularly well, so it's valid but not particularly useful. What does scale (and what's been discussed before by smarter, more eloquent folks) is the impact that good teaching and learning can have on a piece of software's effectiveness.
The impulse to build and use productivity apps is perfectly understandable: productivity is a thinking problem, computers are machines that help you think, so augment yourself with a computer to be more productive. But the productivity apps I've found most useful are incremental improvements of ancient thinking tools: calendars, lists, clocks. The only `new' tool is a browser extension to keep me from doing this:
Computers are indispensable for a few things: sums, remembering, talking to the network. This can make old tools glossier, faster, more convenient as software, but it doesn't make them teach me anything new. By contrast I've never failed to learn something while having a conversation (or just thinking carefully to myself) about my feelings, problems, and approaches with a project.
A brain and a computer make a good team, but two heads are still better. I expect it will stay that way for a long, long time.