One of the common traits I see between talented hackers and app builders is that they steal blocks of time to work on projects whenever they can. For most keen eyed coders those stolen moments are late into the night long past their day jobs and loved ones have all gone to sleep. Spikes in development happen between 10pm-6am, and it's something that should be trackable by doing a little statistical work on public (git) repository activity. I'll do a search when I'm back on the net to see if any work has been done to capture trends in repo action and link to it here.
Early bird's get the worm, but not Traction
With my schedule being 5am-10pm I've got a tough time patterning my development time after the pros. I wake up with my wife and we both head our separate ways by 6-6:15am. Then I walk for 2.5-3hrs while writing a post, read tech docs (RTFM), and finally read up on any interesting startup/tech news by piggy backing wifi from Optimum or open nodes*. I get to my day job by 9:30-10am Mon-Wed/Fri depending on the work load and work till 7-7:30pm. By the time I get home, eat dinner and hang with Michelle a little, it's usually a breath away from bed time and I'm exhausted. Some evenings I can scrape enough energy to go right to work on coding/reading docs from 8:30-11 but it's a fuzzy period (bugs/dumber). For the time being my strongest development days are Thur-Fridays when off, and secondarily Saturday-Sunday afternoons/nights.
Focus
A problem with attempting to squeeze too much into a day is an overall tax on focus. Discipline and will suffer the longer my work days/weeks become. This doesn't matter as much when I'm working on something I'm excited about, but for less challenging tasks, my attention tends to drift and I procrastinate.
A Utopian reality is one in which I can clearly focus on projects of my choosing.
That's my definition of freedom. Couple that focus with a good deal of ownership when it works out, balanced against a zero return (besides learning) if the project sinks. Feeling failure first hand and learning from it at a rapid pace are key to self improvement. Iterate on product design, iterate on application development, even iterate on startups to observe, recognize, and manufacture enormous social and financial trends. We can guess high value opportunities based on current knowledge, but that's only a crude start. We can't know the best product fit without testing and exploring the state of the market. The resultant product design is driven by pressing real world need, coupled with a clear project vision. Building the equivalent of Bacon for the social web and serendipitous search may not be the ideal vision but it's a start ;).
Notes:
*= I've investigated different strategies for developing on my smart phone using dropbox to deploy code but need better shell or terminal support. And the screen real estate is a drag. Still it could open up another hour or two to code. For now I read ebooks, PDFs and instapapered tech docs on the phone while walking to squeeze as much value out of my walking time as I'm able to.