Playtesting is vital. Its hard to know how much is enough, just like re-writing, but certainly some is vital. As a designer its very easy to fall in love with your design, and its even easier to forget its complexity, its fiddly parts and its awkwardness. The first version of SSO marked crew locations with numbers, which seemed fine on paper but absolutely was not, and lasted all of the first playtest when it became colours instead. The most valuable point of playtests is to find out those sorts of human things, which elements are just annoying, which rules people consistently forget, which things aren't obvious and which things get too many questions asked about them.
SSO has a minimum player count of one so the one player version got a hell of a lot of testing, which was one of the reasons I chose it to be our first full project. I have an excellent, smart and accommodating regular gaming group so I get to do a lot of testing with experienced table top gamers. At a playtesting event at Handy Con I had a family herded over to me by the organiser. Now to be honest, I'd been hoping for the sort of group that I could hand the game over to and watch play, running a "blind playtest". What I got were a group entirely unlike any I'd introduced the game to before, who seemed pretty uncertain they even wanted to play it and a good lesson in how stupid I was to want a certain sort of playtester. We had a great game and I got to see people who didn't play co-op games, didn't like sci-fi and I suspect in at least one instance out and out didn't want to play SSO, get taken in and genuinel