It's too early to say if his risks "keep working out". Restructuring is not a risk. His, and others', original decision to make the company a non-profit was also not a calculated risk in this sense.

When he was fired from OpenAI, his use of employee manipulation to regain his position is not a risk; it is the only option he had. It was his bond maturing, of carefully cultivated loyalty he had accrued over years. Gaining that loyalty was not really a risk. It was smart politics.

One risk he took is: signing away such a large portion of the company to Microsoft. I'm not sure whether that is working out.

Another risk he took is: neglecting and sidelining the "safety" portion of his organization. This caused a talent exodus and led to the formation of many competitors. I'm not sure whether that is working out either.