Well, my comments above were meant as a joke. Mostly.
Good article. Of course not all work that I do is the productive, or the so-called 'deliberate practice' kind. As referenced by the fact that I'm following up to my own comments on Hacker News. However an important difference between running a startup and playing a musical instrument is that I'd argue its a lot harder to know what the really effective work is, which means you'll generally have to try a few more things.
No, I'm not quite as bad as the charicature I've described above. I've found that doing a bit of daily exercise is one of the most important things not to give up when working on a startup. You feel and sleep a lot better which affects everything.
I'm working on a startup. My to-do list accumulates tasks faster than I can take items off it. In fact, between starred emails, pen & paper, smartphone apps, web to-dos etc I'm accumulating to-do lists faster than I'm getting through them. I sleep 2 or 3 hours a night. I have no time for friends and my girlfriend is mad at me. I'm in a constant state of stress and anxiety. I can't figure out if I'm on the verge of success or a good for nothing mess.
Do I really have time for AwesomeButUseless.com? Why, of course I do :)
No, the two statements are equivalent. Taking 97% robustness on 2G to mean any given call will drop to be 3/100 then on 3G it is 5/100 which gives a relative probability of 5/3 or 66% more likely.