I think the author is misleading when he says that Python leads to more flow. Flow requires that some conditions hold, one of which is that it is just challenging enough to push conscious thoughts out. I experience flow much more in C++ than in Python, I actually enjoy the complexity. Most people don't get flow from testing because it isn't challenging enough. So what he is doing is lowering the challenge by documenting and removing what he views as "obstacles." He is on the right track, being aware of being in flow, what works for him. I imagine there will be a time when the things he mentions will actually knock him out of flow, and instead he will skip over them without thinking about them.