patience or impatience? Seems to me that being patient would lead to less change, less creep.

Impatience to launch leads to less feature creep.

Seems like he's talking more about patience to stick with your business through "the dip" even if it's not an overnight success rather than patience to first launch. Launch away and iterate, but be patient if it's not an overnight success when you first launch it. It's a good message. Along these lines, I always like to remember that most overnight success stories were 3+ years in the making before the "overnight" success.